mariadb/storage/innobase/include/dict0dict.h

1749 lines
60 KiB
C
Raw Normal View History

/*****************************************************************************
Copyright (c) 1996, 2018, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2012, Facebook Inc.
Copyright (c) 2013, 2022, 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/dict0dict.h
Data dictionary system
Created 1/8/1996 Heikki Tuuri
*******************************************************/
#ifndef dict0dict_h
#define dict0dict_h
#include "data0data.h"
#include "dict0mem.h"
#include "fsp0fsp.h"
#include "srw_lock.h"
#include <my_sys.h>
#include <deque>
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
class MDL_ticket;
MDEV-17794 Do not assign persistent ID for temporary tables InnoDB in MySQL 5.7 introduced two new parameters to the function dict_hdr_get_new_id(), to allow redo logging to be disabled when assigning identifiers to temporary tables or during the backup-unfriendly TRUNCATE TABLE that was replaced in MariaDB by MDEV-13564. Now that MariaDB 10.4.0 removed the crash recovery code for the backup-unfriendly TRUNCATE, we can revert dict_hdr_get_new_id() to be used only for persistent data structures. dict_table_assign_new_id(): Remove. This was a simple 2-line function that was called from few places. dict_table_open_on_id_low(): Declare in the only file where it is called. dict_sys_t::temp_id_hash: A separate lookup table for temporary tables. Table names will be in the common dict_sys_t::table_hash. dict_sys_t::get_temporary_table_id(): Assign a temporary table ID. dict_sys_t::get_table(): Look up a persistent table. dict_sys_t::get_temporary_table(): Look up a temporary table. dict_sys_t::temp_table_id: The sequence of temporary table identifiers. Starts from DICT_HDR_FIRST_ID, so that we can continue to simply compare dict_table_t::id to a few constants for the persistent hard-coded data dictionary tables. undo_node_t::state: Distinguish temporary and persistent tables. lock_check_dict_lock(), lock_get_table_id(): Assert that there cannot be locks on temporary tables. row_rec_to_index_entry_impl(): Assert that there cannot be metadata records on temporary tables. row_undo_ins_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_ins(): Add some assertions. row_undo_mod_clust(), row_undo_mod(): Do not assign node->state. Let row_undo() do that. row_undo_mod_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_try_truncate(): Renamed and simplified from trx_roll_try_truncate(). row_undo_rec_get(): Replaces trx_roll_pop_top_rec_of_trx() and trx_roll_pop_top_rec(). Fetch an undo log record, and assign undo->state accordingly. trx_undo_truncate_end(): Acquire the rseg->mutex only for the minimum required duration, and release it between mini-transactions.
2018-11-22 14:36:50 +01:00
/** the first table or index ID for other than hard-coded system tables */
constexpr uint8_t DICT_HDR_FIRST_ID= 10;
MDEV-17794 Do not assign persistent ID for temporary tables InnoDB in MySQL 5.7 introduced two new parameters to the function dict_hdr_get_new_id(), to allow redo logging to be disabled when assigning identifiers to temporary tables or during the backup-unfriendly TRUNCATE TABLE that was replaced in MariaDB by MDEV-13564. Now that MariaDB 10.4.0 removed the crash recovery code for the backup-unfriendly TRUNCATE, we can revert dict_hdr_get_new_id() to be used only for persistent data structures. dict_table_assign_new_id(): Remove. This was a simple 2-line function that was called from few places. dict_table_open_on_id_low(): Declare in the only file where it is called. dict_sys_t::temp_id_hash: A separate lookup table for temporary tables. Table names will be in the common dict_sys_t::table_hash. dict_sys_t::get_temporary_table_id(): Assign a temporary table ID. dict_sys_t::get_table(): Look up a persistent table. dict_sys_t::get_temporary_table(): Look up a temporary table. dict_sys_t::temp_table_id: The sequence of temporary table identifiers. Starts from DICT_HDR_FIRST_ID, so that we can continue to simply compare dict_table_t::id to a few constants for the persistent hard-coded data dictionary tables. undo_node_t::state: Distinguish temporary and persistent tables. lock_check_dict_lock(), lock_get_table_id(): Assert that there cannot be locks on temporary tables. row_rec_to_index_entry_impl(): Assert that there cannot be metadata records on temporary tables. row_undo_ins_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_ins(): Add some assertions. row_undo_mod_clust(), row_undo_mod(): Do not assign node->state. Let row_undo() do that. row_undo_mod_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_try_truncate(): Renamed and simplified from trx_roll_try_truncate(). row_undo_rec_get(): Replaces trx_roll_pop_top_rec_of_trx() and trx_roll_pop_top_rec(). Fetch an undo log record, and assign undo->state accordingly. trx_undo_truncate_end(): Acquire the rseg->mutex only for the minimum required duration, and release it between mini-transactions.
2018-11-22 14:36:50 +01:00
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
/** Get the database name length in a table name.
@param name filename-safe encoded table name "dbname/tablename"
@return database name length */
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
inline size_t dict_get_db_name_len(const char *name)
{
/* table_name_t::dblen() would assert that '/' is contained */
if (const char* s= strchr(name, '/'))
return size_t(s - name);
return 0;
}
/*********************************************************************//**
Open a table from its database and table name, this is currently used by
foreign constraint parser to get the referenced table.
@return complete table name with database and table name, allocated from
heap memory passed in */
char*
dict_get_referenced_table(
/*======================*/
const char* name, /*!< in: foreign key table name */
const char* database_name, /*!< in: table db name */
ulint database_name_len,/*!< in: db name length */
const char* table_name, /*!< in: table name */
ulint table_name_len, /*!< in: table name length */
dict_table_t** table, /*!< out: table object or NULL */
mem_heap_t* heap, /*!< in: heap memory */
CHARSET_INFO* from_cs); /*!< in: table name charset */
/*********************************************************************//**
Frees a foreign key struct. */
void
dict_foreign_free(
/*==============*/
dict_foreign_t* foreign); /*!< in, own: foreign key struct */
/*********************************************************************//**
Finds the highest [number] for foreign key constraints of the table. Looks
only at the >= 4.0.18-format id's, which are of the form
databasename/tablename_ibfk_[number].
@return highest number, 0 if table has no new format foreign key constraints */
ulint
dict_table_get_highest_foreign_id(
/*==============================*/
dict_table_t* table); /*!< in: table in the dictionary
memory cache */
/** Check whether the dict_table_t is a partition.
A partitioned table on the SQL level is composed of InnoDB tables,
where each InnoDB table is a [sub]partition including its secondary indexes
which belongs to the partition.
@param[in] table Table to check.
@return true if the dict_table_t is a partition else false. */
UNIV_INLINE
bool
dict_table_is_partition(const dict_table_t* table)
{
/* Check both P and p on all platforms in case it was moved to/from
WIN. */
return (strstr(table->name.m_name, "#p#")
|| strstr(table->name.m_name, "#P#"));
}
/********************************************************************//**
Return the end of table name where we have removed dbname and '/'.
@return table name */
const char*
dict_remove_db_name(
/*================*/
const char* name) /*!< in: table name in the form
dbname '/' tablename */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Operation to perform when opening a table */
enum dict_table_op_t {
/** Expect the tablespace to exist. */
DICT_TABLE_OP_NORMAL = 0,
/** Drop any orphan indexes after an aborted online index creation */
DICT_TABLE_OP_DROP_ORPHAN,
/** Silently load the tablespace if it does not exist,
and do not load the definitions of incomplete indexes. */
DICT_TABLE_OP_LOAD_TABLESPACE,
/** Open the table only if it's in table cache. */
DICT_TABLE_OP_OPEN_ONLY_IF_CACHED
};
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
/** Acquire MDL shared for the table name.
@tparam trylock whether to use non-blocking operation
@param[in,out] table table object
@param[in,out] thd background thread
@param[out] mdl mdl ticket
@param[in] table_op operation to perform when opening
@return table object after locking MDL shared
@retval NULL if the table is not readable, or if trylock && MDL blocked */
template<bool trylock, bool purge_thd= false>
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
dict_table_t*
dict_acquire_mdl_shared(dict_table_t *table,
THD *thd,
MDL_ticket **mdl,
dict_table_op_t table_op= DICT_TABLE_OP_NORMAL);
/** Look up a table by numeric identifier.
@param[in] table_id table identifier
@param[in] dict_locked data dictionary locked
@param[in] table_op operation to perform when opening
@param[in,out] thd background thread, or NULL to not acquire MDL
@param[out] mdl mdl ticket, or NULL
@return table, NULL if does not exist */
template<bool purge_thd= false>
dict_table_t*
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
dict_table_open_on_id(table_id_t table_id, bool dict_locked,
dict_table_op_t table_op, THD *thd= nullptr,
MDL_ticket **mdl= nullptr)
MY_ATTRIBUTE((warn_unused_result));
2014-12-22 15:53:17 +01:00
MDEV-25919 preparation: Various cleanup que_eval_sql(): Remove the parameter lock_dict. The only caller with lock_dict=true was dict_stats_exec_sql(), which will now explicitly invoke dict_sys.lock() and dict_sys.unlock() by itself. row_import_cleanup(): Do not unnecessarily lock the dictionary. Concurrent access to the table during ALTER TABLE...IMPORT TABLESPACE is prevented by MDL and the fact that there cannot exist any undo log or change buffer records that would refer to the table or tablespace. row_import_for_mysql(): Do not unnecessarily lock the dictionary while accessing fil_system. Thanks to MDL_EXCLUSIVE that was acquired by the SQL layer, only one IMPORT may be in effect for the table name. row_quiesce_set_state(): Do not unnecessarily lock the dictionary. The dict_table_t::quiesce state is documented to be protected by all index latches, which we are acquiring. dict_table_close(): Introduce a simpler variant with fewer parameters. dict_table_close(): Reduce the amount of calls. We can simply invoke dict_table_t::release() on startup or in DDL operations, or when the table is inaccessible. In none of these cases, there is no need to invalidate the InnoDB persistent statistics. pars_info_t::graph_owns_us: Remove (unused). pars_info_free(): Define inline. fts_delete(), trx_t::evict_table(), row_prebuilt_free(), row_rename_table_for_mysql(): Simplify. row_mysql_lock_data_dictionary(): Remove some references; use dict_sys.lock() and dict_sys.unlock() instead. row_mysql_lock_table(): Remove. Use lock_table_for_trx() instead. ha_innobase::check_if_supported_inplace_alter(), row_create_table_for_mysql(): Simply assert dict_sys.sys_tables_exist(). In commit 49e2c8f0a6fefdeac50925f758090d6bd099768d and commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 srv_start() actually guarantees that the system tables will exist, or the server is in read-only mode, or startup will fail. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:20 +02:00
/** Decrement the count of open handles */
void dict_table_close(dict_table_t *table);
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
/** Decrements the count of open handles of a table.
@param[in,out] table table
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
@param[in] dict_locked whether dict_sys.latch is being held
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
@param[in] thd thread to release MDL
@param[in] mdl metadata lock or NULL if the thread is a
foreground one. */
void
dict_table_close(
MDEV-16678 Prefer MDL to dict_sys.latch for innodb background tasks This is joint work with Thirunarayanan Balathandayuthapani. The MDL interface between InnoDB and the rest of the server (in storage/innobase/dict/dict0dict.cc and in include/) is my work, while most everything else is Thiru's. The collection of InnoDB persistent statistics and the defragmentation were not refactored to use MDL. They will keep relying on lower-level interlocking with fil_check_pending_operations(). The purge of transaction history and the background operations on fulltext indexes will use MDL. We will revert commit 2c4844c9e76427525e8c39a2d72686085efe89c3 (MDEV-17813) because thanks to MDL, purge cannot conflict with DDL operations anymore. For a similar reason, we will remove the MDEV-16222 test case from gcol.innodb_virtual_debug_purge. Purge is essentially replacing all use of the global dict_sys.latch with MDL. Purge will skip the undo log records for tables whose names start with #sql-ib or #sql2. Theoretically, such tables might be renamed back to visible table names if TRUNCATE fails to create a new table, or the final rename in ALTER TABLE...ALGORITHM=COPY fails. In that case, purge could permanently leave some garbage in the table. Such garbage will be tolerated; the table would not be considered corrupted. To avoid repeated MDL releases and acquisitions, trx_purge_attach_undo_recs() will sort undo log records by table_id, and purge_node_t will keep the MDL and table handle open for multiple successive undo log records. get_purge_table(): A new accessor, used during the purge of history for indexed virtual columns. This interface should ideally not exist at all. thd_mdl_context(): Accessor of THD::mdl_context. Wrapped in a new thd_mdl_service. dict_get_db_name_len(): Define inline. dict_acquire_mdl_shared(): Acquire explicit shared MDL on a table name if needed. dict_table_open_on_id(): Return MDL_ticket, if requested. dict_table_close(): Release MDL ticket, if requested. dict_fts_index_syncing(), dict_index_t::index_fts_syncing: Remove. row_drop_table_for_mysql() no longer needs to check these, because MDL guarantees that a fulltext index sync will not be in progress while MDL_EXCLUSIVE is protecting a DDL operation. dict_table_t::parse_name(): Parse the table name for acquiring MDL. purge_node_t::undo_recs: Change the type to std::list<trx_purge_rec_t*> (different container, and storing also roll_ptr). purge_node_t: Add mdl_ticket, last_table_id, purge_thd, mdl_hold_recs for acquiring MDL and for keeping the table open across multiple undo log records. purge_vcol_info_t, row_purge_store_vsec_cur(), row_purge_restore_vsec_cur(): Remove. We will acquire the MDL earlier. purge_sys_t::heap: Added, for reading undo log records. fts_sync_during_ddl(): Invoked during ALGORITHM=INPLACE operations to ensure that fts_sync_table() will not conflict with MDL_EXCLUSIVE. Uses fts_t::sync_message for bookkeeping.
2019-12-10 14:42:50 +01:00
dict_table_t* table,
bool dict_locked,
THD* thd = NULL,
MDL_ticket* mdl = NULL);
/*********************************************************************//**
Gets the minimum number of bytes per character.
@return minimum multi-byte char size, in bytes */
UNIV_INLINE
unsigned
dict_col_get_mbminlen(
/*==================*/
const dict_col_t* col) /*!< in: column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Gets the maximum number of bytes per character.
@return maximum multi-byte char size, in bytes */
UNIV_INLINE
unsigned
dict_col_get_mbmaxlen(
/*==================*/
const dict_col_t* col) /*!< in: column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Gets the column data type. */
UNIV_INLINE
void
dict_col_copy_type(
/*===============*/
const dict_col_t* col, /*!< in: column */
dtype_t* type); /*!< out: data type */
/**********************************************************************//**
Determine bytes of column prefix to be stored in the undo log. Please
note that if !dict_table_has_atomic_blobs(table), no prefix
needs to be stored in the undo log.
@return bytes of column prefix to be stored in the undo log */
UNIV_INLINE
ulint
dict_max_field_len_store_undo(
/*==========================*/
dict_table_t* table, /*!< in: table */
const dict_col_t* col) /*!< in: column which index prefix
is based on */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Determine maximum bytes of a virtual column need to be stored
in the undo log.
@param[in] table dict_table_t for the table
@param[in] col_no virtual column number
@return maximum bytes of virtual column to be stored in the undo log */
UNIV_INLINE
ulint
dict_max_v_field_len_store_undo(
dict_table_t* table,
ulint col_no);
#ifdef UNIV_DEBUG
/*********************************************************************//**
Assert that a column and a data type match.
@return TRUE */
UNIV_INLINE
ibool
dict_col_type_assert_equal(
/*=======================*/
const dict_col_t* col, /*!< in: column */
const dtype_t* type) /*!< in: data type */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
#endif /* UNIV_DEBUG */
/***********************************************************************//**
Returns the minimum size of the column.
@return minimum size */
UNIV_INLINE
unsigned
dict_col_get_min_size(
/*==================*/
const dict_col_t* col) /*!< in: column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/***********************************************************************//**
Returns the maximum size of the column.
@return maximum size */
UNIV_INLINE
ulint
dict_col_get_max_size(
/*==================*/
const dict_col_t* col) /*!< in: column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/***********************************************************************//**
Returns the size of a fixed size column, 0 if not a fixed size column.
@return fixed size, or 0 */
UNIV_INLINE
unsigned
dict_col_get_fixed_size(
/*====================*/
const dict_col_t* col, /*!< in: column */
ulint comp) /*!< in: nonzero=ROW_FORMAT=COMPACT */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/***********************************************************************//**
Returns the ROW_FORMAT=REDUNDANT stored SQL NULL size of a column.
For fixed length types it is the fixed length of the type, otherwise 0.
@return SQL null storage size in ROW_FORMAT=REDUNDANT */
UNIV_INLINE
unsigned
dict_col_get_sql_null_size(
/*=======================*/
const dict_col_t* col, /*!< in: column */
ulint comp) /*!< in: nonzero=ROW_FORMAT=COMPACT */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Gets the column number.
@return col->ind, table column position (starting from 0) */
UNIV_INLINE
unsigned
dict_col_get_no(
/*============*/
const dict_col_t* col) /*!< in: column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Gets the column position in the clustered index. */
UNIV_INLINE
ulint
dict_col_get_clust_pos(
/*===================*/
const dict_col_t* col, /*!< in: table column */
const dict_index_t* clust_index) /*!< in: clustered index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Gets the column position in the given index.
@param[in] col table column
@param[in] index index to be searched for column
@return position of column in the given index. */
UNIV_INLINE
ulint
dict_col_get_index_pos(
const dict_col_t* col,
const dict_index_t* index)
MY_ATTRIBUTE((nonnull, warn_unused_result));
/****************************************************************//**
If the given column name is reserved for InnoDB system columns, return
TRUE.
@return TRUE if name is reserved */
ibool
dict_col_name_is_reserved(
/*======================*/
const char* name) /*!< in: column name */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
MDEV-6076 Persistent AUTO_INCREMENT for InnoDB This should be functionally equivalent to WL#6204 in MySQL 8.0.0, with the notable difference that the file format changes are limited to repurposing a previously unused data field in B-tree pages. For persistent InnoDB tables, write the last used AUTO_INCREMENT value to the root page of the clustered index, in the previously unused (0) PAGE_MAX_TRX_ID field, now aliased as PAGE_ROOT_AUTO_INC. Unlike some other previously unused InnoDB data fields, this one was actually always zero-initialized, at least since MySQL 3.23.49. The writes to PAGE_ROOT_AUTO_INC are protected by SX or X latch on the root page. The SX latch will allow concurrent read access to the root page. (The field PAGE_ROOT_AUTO_INC will only be read on the first-time call to ha_innobase::open() from the SQL layer. The PAGE_ROOT_AUTO_INC can only be updated when executing SQL, so read/write races are not possible.) During INSERT, the PAGE_ROOT_AUTO_INC is updated by the low-level function btr_cur_search_to_nth_level(), adding no extra page access. [Adaptive hash index lookup will be disabled during INSERT.] If some rare UPDATE modifies an AUTO_INCREMENT column, the PAGE_ROOT_AUTO_INC will be adjusted in a separate mini-transaction in ha_innobase::update_row(). When a page is reorganized, we have to preserve the PAGE_ROOT_AUTO_INC field. During ALTER TABLE, the initial AUTO_INCREMENT value will be copied from the table. ALGORITHM=COPY and online log apply in LOCK=NONE will update PAGE_ROOT_AUTO_INC in real time. innodb_col_no(): Determine the dict_table_t::cols[] element index corresponding to a Field of a non-virtual column. (The MySQL 5.7 implementation of virtual columns breaks the 1:1 relationship between Field::field_index and dict_table_t::cols[]. Virtual columns are omitted from dict_table_t::cols[]. Therefore, we must translate the field_index of AUTO_INCREMENT columns into an index of dict_table_t::cols[].) Upgrade from old data files: By default, the AUTO_INCREMENT sequence in old data files would appear to be reset, because PAGE_MAX_TRX_ID or PAGE_ROOT_AUTO_INC would contain the value 0 in each clustered index page. In new data files, PAGE_ROOT_AUTO_INC can only be 0 if the table is empty or does not contain any AUTO_INCREMENT column. For backward compatibility, we use the old method of SELECT MAX(auto_increment_column) for initializing the sequence. btr_read_autoinc(): Read the AUTO_INCREMENT sequence from a new-format data file. btr_read_autoinc_with_fallback(): A variant of btr_read_autoinc() that will resort to reading MAX(auto_increment_column) for data files that did not use AUTO_INCREMENT yet. It was manually tested that during the execution of innodb.autoinc_persist the compatibility logic is not activated (for new files, PAGE_ROOT_AUTO_INC is never 0 in nonempty clustered index root pages). initialize_auto_increment(): Replaces ha_innobase::innobase_initialize_autoinc(). This initializes the AUTO_INCREMENT metadata. Only called from ha_innobase::open(). ha_innobase::info_low(): Do not try to lazily initialize dict_table_t::autoinc. It must already have been initialized by ha_innobase::open() or ha_innobase::create(). Note: The adjustments to class ha_innopart were not tested, because the source code (native InnoDB partitioning) is not being compiled.
2016-12-14 18:56:39 +01:00
/** Unconditionally set the AUTO_INCREMENT counter.
@param[in,out] table table or partition
@param[in] value next available AUTO_INCREMENT value */
MY_ATTRIBUTE((nonnull))
UNIV_INLINE
void
MDEV-6076 Persistent AUTO_INCREMENT for InnoDB This should be functionally equivalent to WL#6204 in MySQL 8.0.0, with the notable difference that the file format changes are limited to repurposing a previously unused data field in B-tree pages. For persistent InnoDB tables, write the last used AUTO_INCREMENT value to the root page of the clustered index, in the previously unused (0) PAGE_MAX_TRX_ID field, now aliased as PAGE_ROOT_AUTO_INC. Unlike some other previously unused InnoDB data fields, this one was actually always zero-initialized, at least since MySQL 3.23.49. The writes to PAGE_ROOT_AUTO_INC are protected by SX or X latch on the root page. The SX latch will allow concurrent read access to the root page. (The field PAGE_ROOT_AUTO_INC will only be read on the first-time call to ha_innobase::open() from the SQL layer. The PAGE_ROOT_AUTO_INC can only be updated when executing SQL, so read/write races are not possible.) During INSERT, the PAGE_ROOT_AUTO_INC is updated by the low-level function btr_cur_search_to_nth_level(), adding no extra page access. [Adaptive hash index lookup will be disabled during INSERT.] If some rare UPDATE modifies an AUTO_INCREMENT column, the PAGE_ROOT_AUTO_INC will be adjusted in a separate mini-transaction in ha_innobase::update_row(). When a page is reorganized, we have to preserve the PAGE_ROOT_AUTO_INC field. During ALTER TABLE, the initial AUTO_INCREMENT value will be copied from the table. ALGORITHM=COPY and online log apply in LOCK=NONE will update PAGE_ROOT_AUTO_INC in real time. innodb_col_no(): Determine the dict_table_t::cols[] element index corresponding to a Field of a non-virtual column. (The MySQL 5.7 implementation of virtual columns breaks the 1:1 relationship between Field::field_index and dict_table_t::cols[]. Virtual columns are omitted from dict_table_t::cols[]. Therefore, we must translate the field_index of AUTO_INCREMENT columns into an index of dict_table_t::cols[].) Upgrade from old data files: By default, the AUTO_INCREMENT sequence in old data files would appear to be reset, because PAGE_MAX_TRX_ID or PAGE_ROOT_AUTO_INC would contain the value 0 in each clustered index page. In new data files, PAGE_ROOT_AUTO_INC can only be 0 if the table is empty or does not contain any AUTO_INCREMENT column. For backward compatibility, we use the old method of SELECT MAX(auto_increment_column) for initializing the sequence. btr_read_autoinc(): Read the AUTO_INCREMENT sequence from a new-format data file. btr_read_autoinc_with_fallback(): A variant of btr_read_autoinc() that will resort to reading MAX(auto_increment_column) for data files that did not use AUTO_INCREMENT yet. It was manually tested that during the execution of innodb.autoinc_persist the compatibility logic is not activated (for new files, PAGE_ROOT_AUTO_INC is never 0 in nonempty clustered index root pages). initialize_auto_increment(): Replaces ha_innobase::innobase_initialize_autoinc(). This initializes the AUTO_INCREMENT metadata. Only called from ha_innobase::open(). ha_innobase::info_low(): Do not try to lazily initialize dict_table_t::autoinc. It must already have been initialized by ha_innobase::open() or ha_innobase::create(). Note: The adjustments to class ha_innopart were not tested, because the source code (native InnoDB partitioning) is not being compiled.
2016-12-14 18:56:39 +01:00
dict_table_autoinc_initialize(dict_table_t* table, ib_uint64_t value)
{
table->autoinc = value;
}
2015-10-09 17:21:46 +02:00
MDEV-6076 Persistent AUTO_INCREMENT for InnoDB This should be functionally equivalent to WL#6204 in MySQL 8.0.0, with the notable difference that the file format changes are limited to repurposing a previously unused data field in B-tree pages. For persistent InnoDB tables, write the last used AUTO_INCREMENT value to the root page of the clustered index, in the previously unused (0) PAGE_MAX_TRX_ID field, now aliased as PAGE_ROOT_AUTO_INC. Unlike some other previously unused InnoDB data fields, this one was actually always zero-initialized, at least since MySQL 3.23.49. The writes to PAGE_ROOT_AUTO_INC are protected by SX or X latch on the root page. The SX latch will allow concurrent read access to the root page. (The field PAGE_ROOT_AUTO_INC will only be read on the first-time call to ha_innobase::open() from the SQL layer. The PAGE_ROOT_AUTO_INC can only be updated when executing SQL, so read/write races are not possible.) During INSERT, the PAGE_ROOT_AUTO_INC is updated by the low-level function btr_cur_search_to_nth_level(), adding no extra page access. [Adaptive hash index lookup will be disabled during INSERT.] If some rare UPDATE modifies an AUTO_INCREMENT column, the PAGE_ROOT_AUTO_INC will be adjusted in a separate mini-transaction in ha_innobase::update_row(). When a page is reorganized, we have to preserve the PAGE_ROOT_AUTO_INC field. During ALTER TABLE, the initial AUTO_INCREMENT value will be copied from the table. ALGORITHM=COPY and online log apply in LOCK=NONE will update PAGE_ROOT_AUTO_INC in real time. innodb_col_no(): Determine the dict_table_t::cols[] element index corresponding to a Field of a non-virtual column. (The MySQL 5.7 implementation of virtual columns breaks the 1:1 relationship between Field::field_index and dict_table_t::cols[]. Virtual columns are omitted from dict_table_t::cols[]. Therefore, we must translate the field_index of AUTO_INCREMENT columns into an index of dict_table_t::cols[].) Upgrade from old data files: By default, the AUTO_INCREMENT sequence in old data files would appear to be reset, because PAGE_MAX_TRX_ID or PAGE_ROOT_AUTO_INC would contain the value 0 in each clustered index page. In new data files, PAGE_ROOT_AUTO_INC can only be 0 if the table is empty or does not contain any AUTO_INCREMENT column. For backward compatibility, we use the old method of SELECT MAX(auto_increment_column) for initializing the sequence. btr_read_autoinc(): Read the AUTO_INCREMENT sequence from a new-format data file. btr_read_autoinc_with_fallback(): A variant of btr_read_autoinc() that will resort to reading MAX(auto_increment_column) for data files that did not use AUTO_INCREMENT yet. It was manually tested that during the execution of innodb.autoinc_persist the compatibility logic is not activated (for new files, PAGE_ROOT_AUTO_INC is never 0 in nonempty clustered index root pages). initialize_auto_increment(): Replaces ha_innobase::innobase_initialize_autoinc(). This initializes the AUTO_INCREMENT metadata. Only called from ha_innobase::open(). ha_innobase::info_low(): Do not try to lazily initialize dict_table_t::autoinc. It must already have been initialized by ha_innobase::open() or ha_innobase::create(). Note: The adjustments to class ha_innopart were not tested, because the source code (native InnoDB partitioning) is not being compiled.
2016-12-14 18:56:39 +01:00
/**
@param[in] table table or partition
@return the next AUTO_INCREMENT counter value
@retval 0 if AUTO_INCREMENT is not yet initialized */
MY_ATTRIBUTE((nonnull, warn_unused_result))
UNIV_INLINE
ib_uint64_t
dict_table_autoinc_read(const dict_table_t* table)
{
return(table->autoinc);
}
2015-10-09 17:21:46 +02:00
MDEV-6076 Persistent AUTO_INCREMENT for InnoDB This should be functionally equivalent to WL#6204 in MySQL 8.0.0, with the notable difference that the file format changes are limited to repurposing a previously unused data field in B-tree pages. For persistent InnoDB tables, write the last used AUTO_INCREMENT value to the root page of the clustered index, in the previously unused (0) PAGE_MAX_TRX_ID field, now aliased as PAGE_ROOT_AUTO_INC. Unlike some other previously unused InnoDB data fields, this one was actually always zero-initialized, at least since MySQL 3.23.49. The writes to PAGE_ROOT_AUTO_INC are protected by SX or X latch on the root page. The SX latch will allow concurrent read access to the root page. (The field PAGE_ROOT_AUTO_INC will only be read on the first-time call to ha_innobase::open() from the SQL layer. The PAGE_ROOT_AUTO_INC can only be updated when executing SQL, so read/write races are not possible.) During INSERT, the PAGE_ROOT_AUTO_INC is updated by the low-level function btr_cur_search_to_nth_level(), adding no extra page access. [Adaptive hash index lookup will be disabled during INSERT.] If some rare UPDATE modifies an AUTO_INCREMENT column, the PAGE_ROOT_AUTO_INC will be adjusted in a separate mini-transaction in ha_innobase::update_row(). When a page is reorganized, we have to preserve the PAGE_ROOT_AUTO_INC field. During ALTER TABLE, the initial AUTO_INCREMENT value will be copied from the table. ALGORITHM=COPY and online log apply in LOCK=NONE will update PAGE_ROOT_AUTO_INC in real time. innodb_col_no(): Determine the dict_table_t::cols[] element index corresponding to a Field of a non-virtual column. (The MySQL 5.7 implementation of virtual columns breaks the 1:1 relationship between Field::field_index and dict_table_t::cols[]. Virtual columns are omitted from dict_table_t::cols[]. Therefore, we must translate the field_index of AUTO_INCREMENT columns into an index of dict_table_t::cols[].) Upgrade from old data files: By default, the AUTO_INCREMENT sequence in old data files would appear to be reset, because PAGE_MAX_TRX_ID or PAGE_ROOT_AUTO_INC would contain the value 0 in each clustered index page. In new data files, PAGE_ROOT_AUTO_INC can only be 0 if the table is empty or does not contain any AUTO_INCREMENT column. For backward compatibility, we use the old method of SELECT MAX(auto_increment_column) for initializing the sequence. btr_read_autoinc(): Read the AUTO_INCREMENT sequence from a new-format data file. btr_read_autoinc_with_fallback(): A variant of btr_read_autoinc() that will resort to reading MAX(auto_increment_column) for data files that did not use AUTO_INCREMENT yet. It was manually tested that during the execution of innodb.autoinc_persist the compatibility logic is not activated (for new files, PAGE_ROOT_AUTO_INC is never 0 in nonempty clustered index root pages). initialize_auto_increment(): Replaces ha_innobase::innobase_initialize_autoinc(). This initializes the AUTO_INCREMENT metadata. Only called from ha_innobase::open(). ha_innobase::info_low(): Do not try to lazily initialize dict_table_t::autoinc. It must already have been initialized by ha_innobase::open() or ha_innobase::create(). Note: The adjustments to class ha_innopart were not tested, because the source code (native InnoDB partitioning) is not being compiled.
2016-12-14 18:56:39 +01:00
/** Update the AUTO_INCREMENT sequence if the value supplied is greater
than the current value.
@param[in,out] table table or partition
@param[in] value AUTO_INCREMENT value that was assigned to a row
@return whether the AUTO_INCREMENT sequence was updated */
MY_ATTRIBUTE((nonnull))
UNIV_INLINE
bool
dict_table_autoinc_update_if_greater(dict_table_t* table, ib_uint64_t value)
{
if (value > table->autoinc) {
table->autoinc = value;
return(true);
}
return(false);
}
/**********************************************************************//**
Adds system columns to a table object. */
void
dict_table_add_system_columns(
/*==========================*/
dict_table_t* table, /*!< in/out: table */
mem_heap_t* heap) /*!< in: temporary heap */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
2015-10-09 17:21:46 +02:00
/**********************************************************************//**
Renames a table object.
@return TRUE if success */
dberr_t
dict_table_rename_in_cache(
/*=======================*/
dict_table_t* table, /*!< in/out: table */
MDEV-13542: Crashing on corrupted page is unhelpful The approach to handling corruption that was chosen by Oracle in commit 177d8b0c125b841c0650d27d735e3b87509dc286 is not really useful. Not only did it actually fail to prevent InnoDB from crashing, but it is making things worse by blocking attempts to rescue data from or rebuild a partially readable table. We will try to prevent crashes in a different way: by propagating errors up the call stack. We will never mark the clustered index persistently corrupted, so that data recovery may be attempted by reading from the table, or by rebuilding the table. This should also fix MDEV-13680 (crash on btr_page_alloc() failure); it was extensively tested with innodb_file_per_table=0 and a non-autoextend system tablespace. We should now avoid crashes in many cases, such as when a page cannot be read or allocated, or an inconsistency is detected when attempting to update multiple pages. We will not crash on double-free, such as on the recovery of DDL in system tablespace in case something was corrupted. Crashes on corrupted data are still possible. The fault injection mechanism that is introduced in the subsequent commit may help catch more of them. buf_page_import_corrupt_failure: Remove the fault injection, and instead corrupt some pages using Perl code in the tests. btr_cur_pessimistic_insert(): Always reserve extents (except for the change buffer), in order to prevent a subsequent allocation failure. btr_pcur_open_at_rnd_pos(): Merged to the only caller ibuf_merge_pages(). btr_assert_not_corrupted(), btr_corruption_report(): Remove. Similar checks are already part of btr_block_get(). FSEG_MAGIC_N_BYTES: Replaces FSEG_MAGIC_N_VALUE. dict_hdr_get(), trx_rsegf_get_new(), trx_undo_page_get(), trx_undo_page_get_s_latched(): Replaced with error-checking calls. trx_rseg_t::get(mtr_t*): Replaces trx_rsegf_get(). trx_rseg_header_create(): Let the caller update the TRX_SYS page if needed. trx_sys_create_sys_pages(): Merged with trx_sysf_create(). dict_check_tablespaces_and_store_max_id(): Do not access DICT_HDR_MAX_SPACE_ID, because it was already recovered in dict_boot(). Merge dict_check_sys_tables() with this function. dir_pathname(): Replaces os_file_make_new_pathname(). row_undo_ins_remove_sec(): Do not modify the undo page by adding a terminating NUL byte to the record. btr_decryption_failed(): Report decryption failures dict_set_corrupted_by_space(), dict_set_encrypted_by_space(), dict_set_corrupted_index_cache_only(): Remove. dict_set_corrupted(): Remove the constant parameter dict_locked=false. Never flag the clustered index corrupted in SYS_INDEXES, because that would deny further access to the table. It might be possible to repair the table by executing ALTER TABLE or OPTIMIZE TABLE, in case no B-tree leaf page is corrupted. dict_table_skip_corrupt_index(), dict_table_next_uncorrupted_index(), row_purge_skip_uncommitted_virtual_index(): Remove, and refactor the callers to read dict_index_t::type only once. dict_table_is_corrupted(): Remove. dict_index_t::is_btree(): Determine if the index is a valid B-tree. BUF_GET_NO_LATCH, BUF_EVICT_IF_IN_POOL: Remove. UNIV_BTR_DEBUG: Remove. Any inconsistency will no longer trigger assertion failures, but error codes being returned. buf_corrupt_page_release(): Replaced with a direct call to buf_pool.corrupted_evict(). fil_invalid_page_access_msg(): Never crash on an invalid read; let the caller of buf_page_get_gen() decide. btr_pcur_t::restore_position(): Propagate failure status to the caller by returning CORRUPTED. opt_search_plan_for_table(): Simplify the code. row_purge_del_mark(), row_purge_upd_exist_or_extern_func(), row_undo_ins_remove_sec_rec(), row_undo_mod_upd_del_sec(), row_undo_mod_del_mark_sec(): Avoid mem_heap_create()/mem_heap_free() when no secondary indexes exist. row_undo_mod_upd_exist_sec(): Simplify the code. row_upd_clust_step(), dict_load_table_one(): Return DB_TABLE_CORRUPT if the clustered index (and therefore the table) is corrupted, similar to what we do in row_insert_for_mysql(). fut_get_ptr(): Replace with buf_page_get_gen() calls. buf_page_get_gen(): Return nullptr and *err=DB_CORRUPTION if the page is marked as freed. For other modes than BUF_GET_POSSIBLY_FREED or BUF_PEEK_IF_IN_POOL this will trigger a debug assertion failure. For BUF_GET_POSSIBLY_FREED, we will return nullptr for freed pages, so that the callers can be simplified. The purge of transaction history will be a new user of BUF_GET_POSSIBLY_FREED, to avoid crashes on corrupted data. buf_page_get_low(): Never crash on a corrupted page, but simply return nullptr. fseg_page_is_allocated(): Replaces fseg_page_is_free(). fts_drop_common_tables(): Return an error if the transaction was rolled back. fil_space_t::set_corrupted(): Report a tablespace as corrupted if it was not reported already. fil_space_t::io(): Invoke fil_space_t::set_corrupted() to report out-of-bounds page access or other errors. Clean up mtr_t::page_lock() buf_page_get_low(): Validate the page identifier (to check for recently read corrupted pages) after acquiring the page latch. buf_page_t::read_complete(): Flag uninitialized (all-zero) pages with DB_FAIL. Return DB_PAGE_CORRUPTED on page number mismatch. mtr_t::defer_drop_ahi(): Renamed from mtr_defer_drop_ahi(). recv_sys_t::free_corrupted_page(): Only set_corrupt_fs() if any log records exist for the page. We do not mind if read-ahead produces corrupted (or all-zero) pages that were not actually needed during recovery. recv_recover_page(): Return whether the operation succeeded. recv_sys_t::recover_low(): Simplify the logic. Check for recovery error. Thanks to Matthias Leich for testing this extensively and to the authors of https://rr-project.org for making it easy to diagnose and fix any failures that were found during the testing.
2022-06-06 13:03:22 +02:00
span<const char> new_name, /*!< in: new name */
bool replace_new_file)
MDEV-17158 TRUNCATE is not atomic after MDEV-13564 It turned out that ha_innobase::truncate() would prematurely commit the transaction already before the completion of the ha_innobase::create(). All of this must be atomic. innodb.truncate_crash: Use the correct DEBUG_SYNC point, and tolerate non-truncation of the table, because the redo log for the TRUNCATE transaction commit might be flushed due to some InnoDB background activity. dict_build_tablespace_for_table(): Merge to the function dict_build_table_def_step(). dict_build_table_def_step(): If a table is being created during an already started data dictionary transaction (such as TRUNCATE), persistently write the table_id to the undo log header before creating any file. In this way, the recovery of TRUNCATE will be able to delete the new file before rolling back the rename of the original table. dict_table_rename_in_cache(): Add the parameter replace_new_file, used as part of rolling back a TRUNCATE operation. fil_rename_tablespace_check(): Add the parameter replace_new. If the parameter is set and a file identified by new_path exists, remove a possible tablespace and also the file. create_table_info_t::create_table_def(): Remove some debug assertions that no longer hold. During TRUNCATE, the transaction will already have been started (and performed a rename operation) before the table is created. Also, remove a call to dict_build_tablespace_for_table(). create_table_info_t::create_table(): Add the parameter create_fk=true. During TRUNCATE TABLE, do not add FOREIGN KEY constraints to the InnoDB data dictionary, because they will also not be removed. row_table_add_foreign_constraints(): If trx=NULL, do not modify the InnoDB data dictionary, but only load the FOREIGN KEY constraints from the data dictionary. ha_innobase::create(): Lock the InnoDB data dictionary cache only if no transaction was passed by the caller. Unlock it in any case. innobase_rename_table(): Add the parameter commit = true. If !commit, do not lock or unlock the data dictionary cache. ha_innobase::truncate(): Lock the data dictionary before invoking rename or create, and let ha_innobase::create() unlock it and also commit or roll back the transaction. trx_undo_mark_as_dict(): Renamed from trx_undo_mark_as_dict_operation() and declared global instead of static. row_undo_ins_parse_undo_rec(): If table_id is set, this must be rolling back the rename operation in TRUNCATE TABLE, and therefore replace_new_file=true.
2018-09-10 13:59:58 +02:00
/*!< in: whether to replace the
file with the new name
(as part of rolling back TRUNCATE) */
MY_ATTRIBUTE((nonnull));
/** Removes an index from the dictionary cache.
@param[in,out] table table whose index to remove
@param[in,out] index index to remove, this object is destroyed and must not
be accessed by the caller afterwards */
void
dict_index_remove_from_cache(
dict_table_t* table,
dict_index_t* index);
/**********************************************************************//**
Change the id of a table object in the dictionary cache. This is used in
DISCARD TABLESPACE. */
void
dict_table_change_id_in_cache(
/*==========================*/
dict_table_t* table, /*!< in/out: table object already in cache */
table_id_t new_id) /*!< in: new id to set */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/**********************************************************************//**
Removes a foreign constraint struct from the dictionary cache. */
void
dict_foreign_remove_from_cache(
/*===========================*/
dict_foreign_t* foreign) /*!< in, own: foreign constraint */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/**********************************************************************//**
Adds a foreign key constraint object to the dictionary cache. May free
the object if there already is an object with the same identifier in.
At least one of foreign table or referenced table must already be in
the dictionary cache!
@return DB_SUCCESS or error code */
dberr_t
dict_foreign_add_to_cache(
/*======================*/
dict_foreign_t* foreign,
/*!< in, own: foreign key constraint */
const char** col_names,
/*!< in: column names, or NULL to use
foreign->foreign_table->col_names */
bool check_charsets,
/*!< in: whether to check charset
compatibility */
dict_err_ignore_t ignore_err)
/*!< in: error to be ignored */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull(1), warn_unused_result));
/**********************************************************************//**
Replace the index passed in with another equivalent index in the
foreign key lists of the table.
@return whether all replacements were found */
bool
dict_foreign_replace_index(
/*=======================*/
dict_table_t* table, /*!< in/out: table */
const char** col_names,
/*!< in: column names, or NULL
to use table->col_names */
const dict_index_t* index) /*!< in: index to be replaced */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull(1,3), warn_unused_result));
/**********************************************************************//**
Parses the CONSTRAINT id's to be dropped in an ALTER TABLE statement.
@return DB_SUCCESS or DB_CANNOT_DROP_CONSTRAINT if syntax error or the
constraint id does not match */
dberr_t
dict_foreign_parse_drop_constraints(
/*================================*/
mem_heap_t* heap, /*!< in: heap from which we can
allocate memory */
trx_t* trx, /*!< in: transaction */
dict_table_t* table, /*!< in: table */
ulint* n, /*!< out: number of constraints
to drop */
const char*** constraints_to_drop) /*!< out: id's of the
constraints to drop */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/**********************************************************************//**
Returns a table object and increments its open handle count.
NOTE! This is a high-level function to be used mainly from outside the
'dict' directory. Inside this directory dict_table_get_low
is usually the appropriate function.
@param[in] table_name Table name
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
@param[in] dict_locked whether dict_sys.latch is being held exclusively
@param[in] ignore_err error to be ignored when loading the table
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
@return table
@retval nullptr if does not exist */
dict_table_t*
dict_table_open_on_name(
const char* table_name,
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
bool dict_locked,
dict_err_ignore_t ignore_err)
MY_ATTRIBUTE((warn_unused_result));
/** Outcome of dict_foreign_find_index() or dict_foreign_qualify_index() */
enum fkerr_t
{
/** A backing index was found for a FOREIGN KEY constraint */
FK_SUCCESS = 0,
/** There is no index that covers the columns in the constraint. */
FK_INDEX_NOT_FOUND,
/** The index is for a prefix index, not a full column. */
FK_IS_PREFIX_INDEX,
/** A condition of SET NULL conflicts with a NOT NULL column. */
FK_COL_NOT_NULL,
/** The column types do not match */
FK_COLS_NOT_EQUAL
};
/*********************************************************************//**
Tries to find an index whose first fields are the columns in the array,
in the same order and is not marked for deletion and is not the same
as types_idx.
@return matching index, NULL if not found */
dict_index_t*
dict_foreign_find_index(
/*====================*/
const dict_table_t* table, /*!< in: table */
const char** col_names,
/*!< in: column names, or NULL
to use table->col_names */
const char** columns,/*!< in: array of column names */
ulint n_cols, /*!< in: number of columns */
const dict_index_t* types_idx,
/*!< in: NULL or an index
whose types the column types
must match */
bool check_charsets,
/*!< in: whether to check
charsets. only has an effect
if types_idx != NULL */
ulint check_null,
/*!< in: nonzero if none of
the columns must be declared
NOT NULL */
fkerr_t* error = NULL, /*!< out: error code */
ulint* err_col_no = NULL,
/*!< out: column number where
error happened */
dict_index_t** err_index = NULL)
/*!< out: index where error
happened */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull(1,3), warn_unused_result));
/** Returns a virtual column's name.
@param[in] table table object
@param[in] col_nr virtual column number(nth virtual column)
@return column name. */
const char*
dict_table_get_v_col_name(
const dict_table_t* table,
ulint col_nr);
/** Check if the table has a given column.
@param[in] table table object
@param[in] col_name column name
@param[in] col_nr column number guessed, 0 as default
@return column number if the table has the specified column,
otherwise table->n_def */
ulint
dict_table_has_column(
const dict_table_t* table,
const char* col_name,
ulint col_nr = 0);
/**********************************************************************//**
Outputs info on foreign keys of a table. */
std::string
dict_print_info_on_foreign_keys(
/*============================*/
ibool create_table_format, /*!< in: if TRUE then print in
a format suitable to be inserted into
a CREATE TABLE, otherwise in the format
of SHOW TABLE STATUS */
trx_t* trx, /*!< in: transaction */
dict_table_t* table); /*!< in: table */
/**********************************************************************//**
Outputs info on a foreign key of a table in a format suitable for
CREATE TABLE. */
std::string
dict_print_info_on_foreign_key_in_create_format(
/*============================================*/
trx_t* trx, /*!< in: transaction */
dict_foreign_t* foreign, /*!< in: foreign key constraint */
ibool add_newline); /*!< in: whether to add a newline */
/*********************************************************************//**
Tries to find an index whose first fields are the columns in the array,
in the same order and is not marked for deletion and is not the same
as types_idx.
@return matching index, NULL if not found */
bool
dict_foreign_qualify_index(
/*====================*/
const dict_table_t* table, /*!< in: table */
const char** col_names,
/*!< in: column names, or NULL
to use table->col_names */
const char** columns,/*!< in: array of column names */
ulint n_cols, /*!< in: number of columns */
const dict_index_t* index, /*!< in: index to check */
const dict_index_t* types_idx,
/*!< in: NULL or an index
whose types the column types
must match */
bool check_charsets,
/*!< in: whether to check
charsets. only has an effect
if types_idx != NULL */
ulint check_null,
/*!< in: nonzero if none of
the columns must be declared
NOT NULL */
fkerr_t* error, /*!< out: error code */
ulint* err_col_no,
/*!< out: column number where
error happened */
dict_index_t** err_index)
/*!< out: index where error
happened */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull(1,3), warn_unused_result));
#ifdef UNIV_DEBUG
/********************************************************************//**
Gets the first index on the table (the clustered index).
@return index, NULL if none exists */
UNIV_INLINE
dict_index_t*
dict_table_get_first_index(
/*=======================*/
const dict_table_t* table) /*!< in: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Gets the last index on the table.
@return index, NULL if none exists */
UNIV_INLINE
dict_index_t*
dict_table_get_last_index(
/*=======================*/
const dict_table_t* table) /*!< in: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Gets the next index on the table.
@return index, NULL if none left */
UNIV_INLINE
dict_index_t*
dict_table_get_next_index(
/*======================*/
const dict_index_t* index) /*!< in: index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
#else /* UNIV_DEBUG */
# define dict_table_get_first_index(table) UT_LIST_GET_FIRST((table)->indexes)
# define dict_table_get_last_index(table) UT_LIST_GET_LAST((table)->indexes)
# define dict_table_get_next_index(index) UT_LIST_GET_NEXT(indexes, index)
#endif /* UNIV_DEBUG */
#define dict_index_is_clust(index) (index)->is_clust()
#define dict_index_is_auto_gen_clust(index) (index)->is_gen_clust()
#define dict_index_is_unique(index) (index)->is_unique()
#define dict_index_is_spatial(index) (index)->is_spatial()
#define dict_index_is_ibuf(index) (index)->is_ibuf()
#define dict_index_is_sec_or_ibuf(index) !(index)->is_primary()
#define dict_index_has_virtual(index) (index)->has_virtual()
/** Get all the FTS indexes on a table.
@param[in] table table
@param[out] indexes all FTS indexes on this table
@return number of FTS indexes */
ulint
dict_table_get_all_fts_indexes(
const dict_table_t* table,
ib_vector_t* indexes);
/********************************************************************//**
Gets the number of user-defined non-virtual columns in a table in the
dictionary cache.
@return number of user-defined (e.g., not ROW_ID) non-virtual
columns of a table */
UNIV_INLINE
unsigned
dict_table_get_n_user_cols(
/*=======================*/
const dict_table_t* table) /*!< in: table */
MY_ATTRIBUTE((warn_unused_result));
/********************************************************************//**
Gets the number of all non-virtual columns (also system) in a table
in the dictionary cache.
@return number of columns of a table */
UNIV_INLINE
unsigned
dict_table_get_n_cols(
/*==================*/
const dict_table_t* table) /*!< in: table */
MY_ATTRIBUTE((warn_unused_result));
/** Gets the number of virtual columns in a table in the dictionary cache.
@param[in] table the table to check
@return number of virtual columns of a table */
UNIV_INLINE
unsigned
dict_table_get_n_v_cols(
const dict_table_t* table);
/** Check if a table has indexed virtual columns
@param[in] table the table to check
@return true is the table has indexed virtual columns */
UNIV_INLINE
bool
dict_table_has_indexed_v_cols(
const dict_table_t* table);
/********************************************************************//**
Gets the approximately estimated number of rows in the table.
@return estimated number of rows */
UNIV_INLINE
ib_uint64_t
dict_table_get_n_rows(
/*==================*/
const dict_table_t* table) /*!< in: table */
MY_ATTRIBUTE((warn_unused_result));
/********************************************************************//**
Increment the number of rows in the table by one.
Notice that this operation is not protected by any latch, the number is
approximate. */
UNIV_INLINE
void
dict_table_n_rows_inc(
/*==================*/
dict_table_t* table) /*!< in/out: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/********************************************************************//**
Decrement the number of rows in the table by one.
Notice that this operation is not protected by any latch, the number is
approximate. */
UNIV_INLINE
void
dict_table_n_rows_dec(
/*==================*/
dict_table_t* table) /*!< in/out: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/** Get nth virtual column
@param[in] table target table
@param[in] col_nr column number in MySQL Table definition
@return dict_v_col_t ptr */
dict_v_col_t*
dict_table_get_nth_v_col_mysql(
const dict_table_t* table,
ulint col_nr);
#ifdef UNIV_DEBUG
/********************************************************************//**
Gets the nth column of a table.
@return pointer to column object */
UNIV_INLINE
dict_col_t*
dict_table_get_nth_col(
/*===================*/
const dict_table_t* table, /*!< in: table */
ulint pos) /*!< in: position of column */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Gets the nth virtual column of a table.
@param[in] table table
@param[in] pos position of virtual column
@return pointer to virtual column object */
UNIV_INLINE
dict_v_col_t*
dict_table_get_nth_v_col(
const dict_table_t* table,
ulint pos);
/********************************************************************//**
Gets the given system column of a table.
@return pointer to column object */
UNIV_INLINE
dict_col_t*
dict_table_get_sys_col(
/*===================*/
const dict_table_t* table, /*!< in: table */
unsigned sys) /*!< in: DATA_ROW_ID, ... */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
#else /* UNIV_DEBUG */
2019-04-03 13:07:18 +02:00
#define dict_table_get_nth_col(table, pos) (&(table)->cols[pos])
#define dict_table_get_sys_col(table, sys) \
&(table)->cols[(table)->n_cols + (sys) - DATA_N_SYS_COLS]
/* Get nth virtual columns */
2019-04-03 13:07:18 +02:00
#define dict_table_get_nth_v_col(table, pos) (&(table)->v_cols[pos])
#endif /* UNIV_DEBUG */
MDEV-11369 Instant ADD COLUMN for InnoDB For InnoDB tables, adding, dropping and reordering columns has required a rebuild of the table and all its indexes. Since MySQL 5.6 (and MariaDB 10.0) this has been supported online (LOCK=NONE), allowing concurrent modification of the tables. This work revises the InnoDB ROW_FORMAT=REDUNDANT, ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC so that columns can be appended instantaneously, with only minor changes performed to the table structure. The counter innodb_instant_alter_column in INFORMATION_SCHEMA.GLOBAL_STATUS is incremented whenever a table rebuild operation is converted into an instant ADD COLUMN operation. ROW_FORMAT=COMPRESSED tables will not support instant ADD COLUMN. Some usability limitations will be addressed in subsequent work: MDEV-13134 Introduce ALTER TABLE attributes ALGORITHM=NOCOPY and ALGORITHM=INSTANT MDEV-14016 Allow instant ADD COLUMN, ADD INDEX, LOCK=NONE The format of the clustered index (PRIMARY KEY) is changed as follows: (1) The FIL_PAGE_TYPE of the root page will be FIL_PAGE_TYPE_INSTANT, and a new field PAGE_INSTANT will contain the original number of fields in the clustered index ('core' fields). If instant ADD COLUMN has not been used or the table becomes empty, or the very first instant ADD COLUMN operation is rolled back, the fields PAGE_INSTANT and FIL_PAGE_TYPE will be reset to 0 and FIL_PAGE_INDEX. (2) A special 'default row' record is inserted into the leftmost leaf, between the page infimum and the first user record. This record is distinguished by the REC_INFO_MIN_REC_FLAG, and it is otherwise in the same format as records that contain values for the instantly added columns. This 'default row' always has the same number of fields as the clustered index according to the table definition. The values of 'core' fields are to be ignored. For other fields, the 'default row' will contain the default values as they were during the ALTER TABLE statement. (If the column default values are changed later, those values will only be stored in the .frm file. The 'default row' will contain the original evaluated values, which must be the same for every row.) The 'default row' must be completely hidden from higher-level access routines. Assertions have been added to ensure that no 'default row' is ever present in the adaptive hash index or in locked records. The 'default row' is never delete-marked. (3) In clustered index leaf page records, the number of fields must reside between the number of 'core' fields (dict_index_t::n_core_fields introduced in this work) and dict_index_t::n_fields. If the number of fields is less than dict_index_t::n_fields, the missing fields are replaced with the column value of the 'default row'. Note: The number of fields in the record may shrink if some of the last instantly added columns are updated to the value that is in the 'default row'. The function btr_cur_trim() implements this 'compression' on update and rollback; dtuple::trim() implements it on insert. (4) In ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC records, the new status value REC_STATUS_COLUMNS_ADDED will indicate the presence of a new record header that will encode n_fields-n_core_fields-1 in 1 or 2 bytes. (In ROW_FORMAT=REDUNDANT records, the record header always explicitly encodes the number of fields.) We introduce the undo log record type TRX_UNDO_INSERT_DEFAULT for covering the insert of the 'default row' record when instant ADD COLUMN is used for the first time. Subsequent instant ADD COLUMN can use TRX_UNDO_UPD_EXIST_REC. This is joint work with Vin Chen (陈福荣) from Tencent. The design that was discussed in April 2017 would not have allowed import or export of data files, because instead of the 'default row' it would have introduced a data dictionary table. The test rpl.rpl_alter_instant is exactly as contributed in pull request #408. The test innodb.instant_alter is based on a contributed test. The redo log record format changes for ROW_FORMAT=DYNAMIC and ROW_FORMAT=COMPACT are as contributed. (With this change present, crash recovery from MariaDB 10.3.1 will fail in spectacular ways!) Also the semantics of higher-level redo log records that modify the PAGE_INSTANT field is changed. The redo log format version identifier was already changed to LOG_HEADER_FORMAT_CURRENT=103 in MariaDB 10.3.1. Everything else has been rewritten by me. Thanks to Elena Stepanova, the code has been tested extensively. When rolling back an instant ADD COLUMN operation, we must empty the PAGE_FREE list after deleting or shortening the 'default row' record, by calling either btr_page_empty() or btr_page_reorganize(). We must know the size of each entry in the PAGE_FREE list. If rollback left a freed copy of the 'default row' in the PAGE_FREE list, we would be unable to determine its size (if it is in ROW_FORMAT=COMPACT or ROW_FORMAT=DYNAMIC) because it would contain more fields than the rolled-back definition of the clustered index. UNIV_SQL_DEFAULT: A new special constant that designates an instantly added column that is not present in the clustered index record. len_is_stored(): Check if a length is an actual length. There are two magic length values: UNIV_SQL_DEFAULT, UNIV_SQL_NULL. dict_col_t::def_val: The 'default row' value of the column. If the column is not added instantly, def_val.len will be UNIV_SQL_DEFAULT. dict_col_t: Add the accessors is_virtual(), is_nullable(), is_instant(), instant_value(). dict_col_t::remove_instant(): Remove the 'instant ADD' status of a column. dict_col_t::name(const dict_table_t& table): Replaces dict_table_get_col_name(). dict_index_t::n_core_fields: The original number of fields. For secondary indexes and if instant ADD COLUMN has not been used, this will be equal to dict_index_t::n_fields. dict_index_t::n_core_null_bytes: Number of bytes needed to represent the null flags; usually equal to UT_BITS_IN_BYTES(n_nullable). dict_index_t::NO_CORE_NULL_BYTES: Magic value signalling that n_core_null_bytes was not initialized yet from the clustered index root page. dict_index_t: Add the accessors is_instant(), is_clust(), get_n_nullable(), instant_field_value(). dict_index_t::instant_add_field(): Adjust clustered index metadata for instant ADD COLUMN. dict_index_t::remove_instant(): Remove the 'instant ADD' status of a clustered index when the table becomes empty, or the very first instant ADD COLUMN operation is rolled back. dict_table_t: Add the accessors is_instant(), is_temporary(), supports_instant(). dict_table_t::instant_add_column(): Adjust metadata for instant ADD COLUMN. dict_table_t::rollback_instant(): Adjust metadata on the rollback of instant ADD COLUMN. prepare_inplace_alter_table_dict(): First create the ctx->new_table, and only then decide if the table really needs to be rebuilt. We must split the creation of table or index metadata from the creation of the dictionary table records and the creation of the data. In this way, we can transform a table-rebuilding operation into an instant ADD COLUMN operation. Dictionary objects will only be added to cache when table rebuilding or index creation is needed. The ctx->instant_table will never be added to cache. dict_table_t::add_to_cache(): Modified and renamed from dict_table_add_to_cache(). Do not modify the table metadata. Let the callers invoke dict_table_add_system_columns() and if needed, set can_be_evicted. dict_create_sys_tables_tuple(), dict_create_table_step(): Omit the system columns (which will now exist in the dict_table_t object already at this point). dict_create_table_step(): Expect the callers to invoke dict_table_add_system_columns(). pars_create_table(): Before creating the table creation execution graph, invoke dict_table_add_system_columns(). row_create_table_for_mysql(): Expect all callers to invoke dict_table_add_system_columns(). create_index_dict(): Replaces row_merge_create_index_graph(). innodb_update_n_cols(): Renamed from innobase_update_n_virtual(). Call my_error() if an error occurs. btr_cur_instant_init(), btr_cur_instant_init_low(), btr_cur_instant_root_init(): Load additional metadata from the clustered index and set dict_index_t::n_core_null_bytes. This is invoked when table metadata is first loaded into the data dictionary. dict_boot(): Initialize n_core_null_bytes for the four hard-coded dictionary tables. dict_create_index_step(): Initialize n_core_null_bytes. This is executed as part of CREATE TABLE. dict_index_build_internal_clust(): Initialize n_core_null_bytes to NO_CORE_NULL_BYTES if table->supports_instant(). row_create_index_for_mysql(): Initialize n_core_null_bytes for CREATE TEMPORARY TABLE. commit_cache_norebuild(): Call the code to rename or enlarge columns in the cache only if instant ADD COLUMN is not being used. (Instant ADD COLUMN would copy all column metadata from instant_table to old_table, including the names and lengths.) PAGE_INSTANT: A new 13-bit field for storing dict_index_t::n_core_fields. This is repurposing the 16-bit field PAGE_DIRECTION, of which only the least significant 3 bits were used. The original byte containing PAGE_DIRECTION will be accessible via the new constant PAGE_DIRECTION_B. page_get_instant(), page_set_instant(): Accessors for the PAGE_INSTANT. page_ptr_get_direction(), page_get_direction(), page_ptr_set_direction(): Accessors for PAGE_DIRECTION. page_direction_reset(): Reset PAGE_DIRECTION, PAGE_N_DIRECTION. page_direction_increment(): Increment PAGE_N_DIRECTION and set PAGE_DIRECTION. rec_get_offsets(): Use the 'leaf' parameter for non-debug purposes, and assume that heap_no is always set. Initialize all dict_index_t::n_fields for ROW_FORMAT=REDUNDANT records, even if the record contains fewer fields. rec_offs_make_valid(): Add the parameter 'leaf'. rec_copy_prefix_to_dtuple(): Assert that the tuple is only built on the core fields. Instant ADD COLUMN only applies to the clustered index, and we should never build a search key that has more than the PRIMARY KEY and possibly DB_TRX_ID,DB_ROLL_PTR. All these columns are always present. dict_index_build_data_tuple(): Remove assertions that would be duplicated in rec_copy_prefix_to_dtuple(). rec_init_offsets(): Support ROW_FORMAT=REDUNDANT records whose number of fields is between n_core_fields and n_fields. cmp_rec_rec_with_match(): Implement the comparison between two MIN_REC_FLAG records. trx_t::in_rollback: Make the field available in non-debug builds. trx_start_for_ddl_low(): Remove dangerous error-tolerance. A dictionary transaction must be flagged as such before it has generated any undo log records. This is because trx_undo_assign_undo() will mark the transaction as a dictionary transaction in the undo log header right before the very first undo log record is being written. btr_index_rec_validate(): Account for instant ADD COLUMN row_undo_ins_remove_clust_rec(): On the rollback of an insert into SYS_COLUMNS, revert instant ADD COLUMN in the cache by removing the last column from the table and the clustered index. row_search_on_row_ref(), row_undo_mod_parse_undo_rec(), row_undo_mod(), trx_undo_update_rec_get_update(): Handle the 'default row' as a special case. dtuple_t::trim(index): Omit a redundant suffix of an index tuple right before insert or update. After instant ADD COLUMN, if the last fields of a clustered index tuple match the 'default row', there is no need to store them. While trimming the entry, we must hold a page latch, so that the table cannot be emptied and the 'default row' be deleted. btr_cur_optimistic_update(), btr_cur_pessimistic_update(), row_upd_clust_rec_by_insert(), row_ins_clust_index_entry_low(): Invoke dtuple_t::trim() if needed. row_ins_clust_index_entry(): Restore dtuple_t::n_fields after calling row_ins_clust_index_entry_low(). rec_get_converted_size(), rec_get_converted_size_comp(): Allow the number of fields to be between n_core_fields and n_fields. Do not support infimum,supremum. They are never supposed to be stored in dtuple_t, because page creation nowadays uses a lower-level method for initializing them. rec_convert_dtuple_to_rec_comp(): Assign the status bits based on the number of fields. btr_cur_trim(): In an update, trim the index entry as needed. For the 'default row', handle rollback specially. For user records, omit fields that match the 'default row'. btr_cur_optimistic_delete_func(), btr_cur_pessimistic_delete(): Skip locking and adaptive hash index for the 'default row'. row_log_table_apply_convert_mrec(): Replace 'default row' values if needed. In the temporary file that is applied by row_log_table_apply(), we must identify whether the records contain the extra header for instantly added columns. For now, we will allocate an additional byte for this for ROW_T_INSERT and ROW_T_UPDATE records when the source table has been subject to instant ADD COLUMN. The ROW_T_DELETE records are fine, as they will be converted and will only contain 'core' columns (PRIMARY KEY and some system columns) that are converted from dtuple_t. rec_get_converted_size_temp(), rec_init_offsets_temp(), rec_convert_dtuple_to_temp(): Add the parameter 'status'. REC_INFO_DEFAULT_ROW = REC_INFO_MIN_REC_FLAG | REC_STATUS_COLUMNS_ADDED: An info_bits constant for distinguishing the 'default row' record. rec_comp_status_t: An enum of the status bit values. rec_leaf_format: An enum that replaces the bool parameter of rec_init_offsets_comp_ordinary().
2017-10-06 06:00:05 +02:00
/** Wrapper function.
@see dict_col_t::name()
@param[in] table table
@param[in] col_nr column number in table
@return column name */
inline
const char*
dict_table_get_col_name(const dict_table_t* table, ulint col_nr)
{
return(dict_table_get_nth_col(table, col_nr)->name(*table));
}
/********************************************************************//**
Gets the given system column number of a table.
@return column number */
UNIV_INLINE
unsigned
dict_table_get_sys_col_no(
/*======================*/
const dict_table_t* table, /*!< in: table */
unsigned sys) /*!< in: DATA_ROW_ID, ... */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Returns the minimum data size of an index record.
@return minimum data size in bytes */
UNIV_INLINE
unsigned
dict_index_get_min_size(
/*====================*/
const dict_index_t* index) /*!< in: index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
2018-11-12 16:01:56 +01:00
#define dict_table_is_comp(table) (table)->not_redundant()
/** Determine if a table uses atomic BLOBs (no locally stored prefix).
@param[in] table InnoDB table
@return whether BLOBs are atomic */
inline
bool
dict_table_has_atomic_blobs(const dict_table_t* table)
{
return(DICT_TF_HAS_ATOMIC_BLOBS(table->flags));
}
2019-07-25 17:42:06 +02:00
/** @return potential max length stored inline for externally stored fields */
inline size_t dict_table_t::get_overflow_field_local_len() const
{
if (dict_table_has_atomic_blobs(this)) {
/* ROW_FORMAT=DYNAMIC or ROW_FORMAT=COMPRESSED: do not
store any BLOB prefix locally */
return BTR_EXTERN_FIELD_REF_SIZE;
}
/* up to MySQL 5.1: store a 768-byte prefix locally */
return BTR_EXTERN_FIELD_REF_SIZE + DICT_ANTELOPE_MAX_INDEX_COL_LEN;
}
/** Set the various values in a dict_table_t::flags pointer.
@param[in,out] flags, Pointer to a 4 byte Table Flags
@param[in] format, File Format
@param[in] zip_ssize Zip Shift Size
@param[in] use_data_dir Table uses DATA DIRECTORY
@param[in] page_compressed Table uses page compression
@param[in] page_compression_level Page compression level */
UNIV_INLINE
void
dict_tf_set(
ulint* flags,
rec_format_t format,
ulint zip_ssize,
bool use_data_dir,
bool page_compressed,
ulint page_compression_level);
/** Convert a 32 bit integer table flags to the 32 bit FSP Flags.
Fsp Flags are written into the tablespace header at the offset
FSP_SPACE_FLAGS and are also stored in the fil_space_t::flags field.
The following chart shows the translation of the low order bit.
Other bits are the same.
========================= Low order bit ==========================
| REDUNDANT | COMPACT | COMPRESSED | DYNAMIC
dict_table_t::flags | 0 | 1 | 1 | 1
fil_space_t::flags | 0 | 0 | 1 | 1
==================================================================
@param[in] table_flags dict_table_t::flags
@return tablespace flags (fil_space_t::flags) */
UNIV_INLINE
ulint
dict_tf_to_fsp_flags(ulint table_flags)
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((const));
/** Extract the ROW_FORMAT=COMPRESSED page size from table flags.
@param[in] flags flags
@return ROW_FORMAT=COMPRESSED page size
@retval 0 if not compressed */
inline ulint dict_tf_get_zip_size(ulint flags)
{
flags &= DICT_TF_MASK_ZIP_SSIZE;
return flags
? (UNIV_ZIP_SIZE_MIN >> 1)
<< (FSP_FLAGS_GET_ZIP_SSIZE(flags >> DICT_TF_POS_ZIP_SSIZE
<< FSP_FLAGS_POS_ZIP_SSIZE))
: 0;
}
/********************************************************************//**
Checks if a column is in the ordering columns of the clustered index of a
table. Column prefixes are treated like whole columns.
@return TRUE if the column, or its prefix, is in the clustered key */
ibool
dict_table_col_in_clustered_key(
/*============================*/
const dict_table_t* table, /*!< in: table */
ulint n) /*!< in: column number */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*******************************************************************//**
Check if the table has an FTS index.
@return TRUE if table has an FTS index */
UNIV_INLINE
ibool
dict_table_has_fts_index(
/*=====================*/
dict_table_t* table) /*!< in: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Copies types of virtual columns contained in table to tuple and sets all
fields of the tuple to the SQL NULL value. This function should
be called right after dtuple_create().
@param[in,out] tuple data tuple
@param[in] table table
*/
void
dict_table_copy_v_types(
dtuple_t* tuple,
const dict_table_t* table);
/*******************************************************************//**
Copies types of columns contained in table to tuple and sets all
fields of the tuple to the SQL NULL value. This function should
be called right after dtuple_create(). */
void
dict_table_copy_types(
/*==================*/
dtuple_t* tuple, /*!< in/out: data tuple */
const dict_table_t* table) /*!< in: table */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/** Adds an index to the dictionary cache, with possible indexing newly
added column.
@param[in,out] index index; NOTE! The index memory
object is freed in this function!
@param[in] page_no root page number of the index
@param[in] add_v virtual columns being added along with ADD INDEX
@return DB_SUCCESS, or DB_CORRUPTION */
dberr_t
dict_index_add_to_cache(
dict_index_t*& index,
ulint page_no,
const dict_add_v_col_t* add_v = NULL)
MY_ATTRIBUTE((warn_unused_result));
/********************************************************************//**
Gets the number of fields in the internal representation of an index,
including fields added by the dictionary system.
@return number of fields */
UNIV_INLINE
MDEV-21907: InnoDB: Enable -Wconversion on clang and GCC The -Wconversion in GCC seems to be stricter than in clang. GCC at least since version 4.4.7 issues truncation warnings for assignments to bitfields, while clang 10 appears to only issue warnings when the sizes in bytes rounded to the nearest integer powers of 2 are different. Before GCC 10.0.0, -Wconversion required more casts and would not allow some operations, such as x<<=1 or x+=1 on a data type that is narrower than int. GCC 5 (but not GCC 4, GCC 6, or any later version) is complaining about x|=y even when x and y are compatible types that are narrower than int. Hence, we must rewrite some x|=y as x=static_cast<byte>(x|y) or similar, or we must disable -Wconversion. In GCC 6 and later, the warning for assigning wider to bitfields that are narrower than 8, 16, or 32 bits can be suppressed by applying a bitwise & with the exact bitmask of the bitfield. For older GCC, we must disable -Wconversion for GCC 4 or 5 in such cases. The bitwise negation operator appears to promote short integers to a wider type, and hence we must add explicit truncation casts around them. Microsoft Visual C does not allow a static_cast to truncate a constant, such as static_cast<byte>(1) truncating int. Hence, we will use the constructor-style cast byte(~1) for such cases. This has been tested at least with GCC 4.8.5, 5.4.0, 7.4.0, 9.2.1, 10.0.0, clang 9.0.1, 10.0.0, and MSVC 14.22.27905 (Microsoft Visual Studio 2019) on 64-bit and 32-bit targets (IA-32, AMD64, POWER 8, POWER 9, ARMv8).
2020-03-12 18:46:41 +01:00
uint16_t
dict_index_get_n_fields(
/*====================*/
const dict_index_t* index) /*!< in: an internal
representation of index (in
the dictionary cache) */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
MDEV-11369 Instant ADD COLUMN for InnoDB For InnoDB tables, adding, dropping and reordering columns has required a rebuild of the table and all its indexes. Since MySQL 5.6 (and MariaDB 10.0) this has been supported online (LOCK=NONE), allowing concurrent modification of the tables. This work revises the InnoDB ROW_FORMAT=REDUNDANT, ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC so that columns can be appended instantaneously, with only minor changes performed to the table structure. The counter innodb_instant_alter_column in INFORMATION_SCHEMA.GLOBAL_STATUS is incremented whenever a table rebuild operation is converted into an instant ADD COLUMN operation. ROW_FORMAT=COMPRESSED tables will not support instant ADD COLUMN. Some usability limitations will be addressed in subsequent work: MDEV-13134 Introduce ALTER TABLE attributes ALGORITHM=NOCOPY and ALGORITHM=INSTANT MDEV-14016 Allow instant ADD COLUMN, ADD INDEX, LOCK=NONE The format of the clustered index (PRIMARY KEY) is changed as follows: (1) The FIL_PAGE_TYPE of the root page will be FIL_PAGE_TYPE_INSTANT, and a new field PAGE_INSTANT will contain the original number of fields in the clustered index ('core' fields). If instant ADD COLUMN has not been used or the table becomes empty, or the very first instant ADD COLUMN operation is rolled back, the fields PAGE_INSTANT and FIL_PAGE_TYPE will be reset to 0 and FIL_PAGE_INDEX. (2) A special 'default row' record is inserted into the leftmost leaf, between the page infimum and the first user record. This record is distinguished by the REC_INFO_MIN_REC_FLAG, and it is otherwise in the same format as records that contain values for the instantly added columns. This 'default row' always has the same number of fields as the clustered index according to the table definition. The values of 'core' fields are to be ignored. For other fields, the 'default row' will contain the default values as they were during the ALTER TABLE statement. (If the column default values are changed later, those values will only be stored in the .frm file. The 'default row' will contain the original evaluated values, which must be the same for every row.) The 'default row' must be completely hidden from higher-level access routines. Assertions have been added to ensure that no 'default row' is ever present in the adaptive hash index or in locked records. The 'default row' is never delete-marked. (3) In clustered index leaf page records, the number of fields must reside between the number of 'core' fields (dict_index_t::n_core_fields introduced in this work) and dict_index_t::n_fields. If the number of fields is less than dict_index_t::n_fields, the missing fields are replaced with the column value of the 'default row'. Note: The number of fields in the record may shrink if some of the last instantly added columns are updated to the value that is in the 'default row'. The function btr_cur_trim() implements this 'compression' on update and rollback; dtuple::trim() implements it on insert. (4) In ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC records, the new status value REC_STATUS_COLUMNS_ADDED will indicate the presence of a new record header that will encode n_fields-n_core_fields-1 in 1 or 2 bytes. (In ROW_FORMAT=REDUNDANT records, the record header always explicitly encodes the number of fields.) We introduce the undo log record type TRX_UNDO_INSERT_DEFAULT for covering the insert of the 'default row' record when instant ADD COLUMN is used for the first time. Subsequent instant ADD COLUMN can use TRX_UNDO_UPD_EXIST_REC. This is joint work with Vin Chen (陈福荣) from Tencent. The design that was discussed in April 2017 would not have allowed import or export of data files, because instead of the 'default row' it would have introduced a data dictionary table. The test rpl.rpl_alter_instant is exactly as contributed in pull request #408. The test innodb.instant_alter is based on a contributed test. The redo log record format changes for ROW_FORMAT=DYNAMIC and ROW_FORMAT=COMPACT are as contributed. (With this change present, crash recovery from MariaDB 10.3.1 will fail in spectacular ways!) Also the semantics of higher-level redo log records that modify the PAGE_INSTANT field is changed. The redo log format version identifier was already changed to LOG_HEADER_FORMAT_CURRENT=103 in MariaDB 10.3.1. Everything else has been rewritten by me. Thanks to Elena Stepanova, the code has been tested extensively. When rolling back an instant ADD COLUMN operation, we must empty the PAGE_FREE list after deleting or shortening the 'default row' record, by calling either btr_page_empty() or btr_page_reorganize(). We must know the size of each entry in the PAGE_FREE list. If rollback left a freed copy of the 'default row' in the PAGE_FREE list, we would be unable to determine its size (if it is in ROW_FORMAT=COMPACT or ROW_FORMAT=DYNAMIC) because it would contain more fields than the rolled-back definition of the clustered index. UNIV_SQL_DEFAULT: A new special constant that designates an instantly added column that is not present in the clustered index record. len_is_stored(): Check if a length is an actual length. There are two magic length values: UNIV_SQL_DEFAULT, UNIV_SQL_NULL. dict_col_t::def_val: The 'default row' value of the column. If the column is not added instantly, def_val.len will be UNIV_SQL_DEFAULT. dict_col_t: Add the accessors is_virtual(), is_nullable(), is_instant(), instant_value(). dict_col_t::remove_instant(): Remove the 'instant ADD' status of a column. dict_col_t::name(const dict_table_t& table): Replaces dict_table_get_col_name(). dict_index_t::n_core_fields: The original number of fields. For secondary indexes and if instant ADD COLUMN has not been used, this will be equal to dict_index_t::n_fields. dict_index_t::n_core_null_bytes: Number of bytes needed to represent the null flags; usually equal to UT_BITS_IN_BYTES(n_nullable). dict_index_t::NO_CORE_NULL_BYTES: Magic value signalling that n_core_null_bytes was not initialized yet from the clustered index root page. dict_index_t: Add the accessors is_instant(), is_clust(), get_n_nullable(), instant_field_value(). dict_index_t::instant_add_field(): Adjust clustered index metadata for instant ADD COLUMN. dict_index_t::remove_instant(): Remove the 'instant ADD' status of a clustered index when the table becomes empty, or the very first instant ADD COLUMN operation is rolled back. dict_table_t: Add the accessors is_instant(), is_temporary(), supports_instant(). dict_table_t::instant_add_column(): Adjust metadata for instant ADD COLUMN. dict_table_t::rollback_instant(): Adjust metadata on the rollback of instant ADD COLUMN. prepare_inplace_alter_table_dict(): First create the ctx->new_table, and only then decide if the table really needs to be rebuilt. We must split the creation of table or index metadata from the creation of the dictionary table records and the creation of the data. In this way, we can transform a table-rebuilding operation into an instant ADD COLUMN operation. Dictionary objects will only be added to cache when table rebuilding or index creation is needed. The ctx->instant_table will never be added to cache. dict_table_t::add_to_cache(): Modified and renamed from dict_table_add_to_cache(). Do not modify the table metadata. Let the callers invoke dict_table_add_system_columns() and if needed, set can_be_evicted. dict_create_sys_tables_tuple(), dict_create_table_step(): Omit the system columns (which will now exist in the dict_table_t object already at this point). dict_create_table_step(): Expect the callers to invoke dict_table_add_system_columns(). pars_create_table(): Before creating the table creation execution graph, invoke dict_table_add_system_columns(). row_create_table_for_mysql(): Expect all callers to invoke dict_table_add_system_columns(). create_index_dict(): Replaces row_merge_create_index_graph(). innodb_update_n_cols(): Renamed from innobase_update_n_virtual(). Call my_error() if an error occurs. btr_cur_instant_init(), btr_cur_instant_init_low(), btr_cur_instant_root_init(): Load additional metadata from the clustered index and set dict_index_t::n_core_null_bytes. This is invoked when table metadata is first loaded into the data dictionary. dict_boot(): Initialize n_core_null_bytes for the four hard-coded dictionary tables. dict_create_index_step(): Initialize n_core_null_bytes. This is executed as part of CREATE TABLE. dict_index_build_internal_clust(): Initialize n_core_null_bytes to NO_CORE_NULL_BYTES if table->supports_instant(). row_create_index_for_mysql(): Initialize n_core_null_bytes for CREATE TEMPORARY TABLE. commit_cache_norebuild(): Call the code to rename or enlarge columns in the cache only if instant ADD COLUMN is not being used. (Instant ADD COLUMN would copy all column metadata from instant_table to old_table, including the names and lengths.) PAGE_INSTANT: A new 13-bit field for storing dict_index_t::n_core_fields. This is repurposing the 16-bit field PAGE_DIRECTION, of which only the least significant 3 bits were used. The original byte containing PAGE_DIRECTION will be accessible via the new constant PAGE_DIRECTION_B. page_get_instant(), page_set_instant(): Accessors for the PAGE_INSTANT. page_ptr_get_direction(), page_get_direction(), page_ptr_set_direction(): Accessors for PAGE_DIRECTION. page_direction_reset(): Reset PAGE_DIRECTION, PAGE_N_DIRECTION. page_direction_increment(): Increment PAGE_N_DIRECTION and set PAGE_DIRECTION. rec_get_offsets(): Use the 'leaf' parameter for non-debug purposes, and assume that heap_no is always set. Initialize all dict_index_t::n_fields for ROW_FORMAT=REDUNDANT records, even if the record contains fewer fields. rec_offs_make_valid(): Add the parameter 'leaf'. rec_copy_prefix_to_dtuple(): Assert that the tuple is only built on the core fields. Instant ADD COLUMN only applies to the clustered index, and we should never build a search key that has more than the PRIMARY KEY and possibly DB_TRX_ID,DB_ROLL_PTR. All these columns are always present. dict_index_build_data_tuple(): Remove assertions that would be duplicated in rec_copy_prefix_to_dtuple(). rec_init_offsets(): Support ROW_FORMAT=REDUNDANT records whose number of fields is between n_core_fields and n_fields. cmp_rec_rec_with_match(): Implement the comparison between two MIN_REC_FLAG records. trx_t::in_rollback: Make the field available in non-debug builds. trx_start_for_ddl_low(): Remove dangerous error-tolerance. A dictionary transaction must be flagged as such before it has generated any undo log records. This is because trx_undo_assign_undo() will mark the transaction as a dictionary transaction in the undo log header right before the very first undo log record is being written. btr_index_rec_validate(): Account for instant ADD COLUMN row_undo_ins_remove_clust_rec(): On the rollback of an insert into SYS_COLUMNS, revert instant ADD COLUMN in the cache by removing the last column from the table and the clustered index. row_search_on_row_ref(), row_undo_mod_parse_undo_rec(), row_undo_mod(), trx_undo_update_rec_get_update(): Handle the 'default row' as a special case. dtuple_t::trim(index): Omit a redundant suffix of an index tuple right before insert or update. After instant ADD COLUMN, if the last fields of a clustered index tuple match the 'default row', there is no need to store them. While trimming the entry, we must hold a page latch, so that the table cannot be emptied and the 'default row' be deleted. btr_cur_optimistic_update(), btr_cur_pessimistic_update(), row_upd_clust_rec_by_insert(), row_ins_clust_index_entry_low(): Invoke dtuple_t::trim() if needed. row_ins_clust_index_entry(): Restore dtuple_t::n_fields after calling row_ins_clust_index_entry_low(). rec_get_converted_size(), rec_get_converted_size_comp(): Allow the number of fields to be between n_core_fields and n_fields. Do not support infimum,supremum. They are never supposed to be stored in dtuple_t, because page creation nowadays uses a lower-level method for initializing them. rec_convert_dtuple_to_rec_comp(): Assign the status bits based on the number of fields. btr_cur_trim(): In an update, trim the index entry as needed. For the 'default row', handle rollback specially. For user records, omit fields that match the 'default row'. btr_cur_optimistic_delete_func(), btr_cur_pessimistic_delete(): Skip locking and adaptive hash index for the 'default row'. row_log_table_apply_convert_mrec(): Replace 'default row' values if needed. In the temporary file that is applied by row_log_table_apply(), we must identify whether the records contain the extra header for instantly added columns. For now, we will allocate an additional byte for this for ROW_T_INSERT and ROW_T_UPDATE records when the source table has been subject to instant ADD COLUMN. The ROW_T_DELETE records are fine, as they will be converted and will only contain 'core' columns (PRIMARY KEY and some system columns) that are converted from dtuple_t. rec_get_converted_size_temp(), rec_init_offsets_temp(), rec_convert_dtuple_to_temp(): Add the parameter 'status'. REC_INFO_DEFAULT_ROW = REC_INFO_MIN_REC_FLAG | REC_STATUS_COLUMNS_ADDED: An info_bits constant for distinguishing the 'default row' record. rec_comp_status_t: An enum of the status bit values. rec_leaf_format: An enum that replaces the bool parameter of rec_init_offsets_comp_ordinary().
2017-10-06 06:00:05 +02:00
/********************************************************************//**
Gets the number of fields in the internal representation of an index
that uniquely determine the position of an index entry in the index, if
we do not take multiversioning into account: in the B-tree use the value
returned by dict_index_get_n_unique_in_tree.
@return number of fields */
UNIV_INLINE
MDEV-21907: InnoDB: Enable -Wconversion on clang and GCC The -Wconversion in GCC seems to be stricter than in clang. GCC at least since version 4.4.7 issues truncation warnings for assignments to bitfields, while clang 10 appears to only issue warnings when the sizes in bytes rounded to the nearest integer powers of 2 are different. Before GCC 10.0.0, -Wconversion required more casts and would not allow some operations, such as x<<=1 or x+=1 on a data type that is narrower than int. GCC 5 (but not GCC 4, GCC 6, or any later version) is complaining about x|=y even when x and y are compatible types that are narrower than int. Hence, we must rewrite some x|=y as x=static_cast<byte>(x|y) or similar, or we must disable -Wconversion. In GCC 6 and later, the warning for assigning wider to bitfields that are narrower than 8, 16, or 32 bits can be suppressed by applying a bitwise & with the exact bitmask of the bitfield. For older GCC, we must disable -Wconversion for GCC 4 or 5 in such cases. The bitwise negation operator appears to promote short integers to a wider type, and hence we must add explicit truncation casts around them. Microsoft Visual C does not allow a static_cast to truncate a constant, such as static_cast<byte>(1) truncating int. Hence, we will use the constructor-style cast byte(~1) for such cases. This has been tested at least with GCC 4.8.5, 5.4.0, 7.4.0, 9.2.1, 10.0.0, clang 9.0.1, 10.0.0, and MSVC 14.22.27905 (Microsoft Visual Studio 2019) on 64-bit and 32-bit targets (IA-32, AMD64, POWER 8, POWER 9, ARMv8).
2020-03-12 18:46:41 +01:00
uint16_t
dict_index_get_n_unique(
/*====================*/
const dict_index_t* index) /*!< in: an internal representation
of index (in the dictionary cache) */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Gets the number of fields in the internal representation of an index
which uniquely determine the position of an index entry in the index, if
we also take multiversioning into account.
@return number of fields */
UNIV_INLINE
MDEV-21907: InnoDB: Enable -Wconversion on clang and GCC The -Wconversion in GCC seems to be stricter than in clang. GCC at least since version 4.4.7 issues truncation warnings for assignments to bitfields, while clang 10 appears to only issue warnings when the sizes in bytes rounded to the nearest integer powers of 2 are different. Before GCC 10.0.0, -Wconversion required more casts and would not allow some operations, such as x<<=1 or x+=1 on a data type that is narrower than int. GCC 5 (but not GCC 4, GCC 6, or any later version) is complaining about x|=y even when x and y are compatible types that are narrower than int. Hence, we must rewrite some x|=y as x=static_cast<byte>(x|y) or similar, or we must disable -Wconversion. In GCC 6 and later, the warning for assigning wider to bitfields that are narrower than 8, 16, or 32 bits can be suppressed by applying a bitwise & with the exact bitmask of the bitfield. For older GCC, we must disable -Wconversion for GCC 4 or 5 in such cases. The bitwise negation operator appears to promote short integers to a wider type, and hence we must add explicit truncation casts around them. Microsoft Visual C does not allow a static_cast to truncate a constant, such as static_cast<byte>(1) truncating int. Hence, we will use the constructor-style cast byte(~1) for such cases. This has been tested at least with GCC 4.8.5, 5.4.0, 7.4.0, 9.2.1, 10.0.0, clang 9.0.1, 10.0.0, and MSVC 14.22.27905 (Microsoft Visual Studio 2019) on 64-bit and 32-bit targets (IA-32, AMD64, POWER 8, POWER 9, ARMv8).
2020-03-12 18:46:41 +01:00
uint16_t
dict_index_get_n_unique_in_tree(
/*============================*/
const dict_index_t* index) /*!< in: an internal representation
of index (in the dictionary cache) */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** The number of fields in the nonleaf page of spatial index, except
the page no field. */
#define DICT_INDEX_SPATIAL_NODEPTR_SIZE 1
/**
Gets the number of fields on nonleaf page level in the internal representation
of an index which uniquely determine the position of an index entry in the
index, if we also take multiversioning into account. Note, it doesn't
include page no field.
@param[in] index index
@return number of fields */
UNIV_INLINE
MDEV-21907: InnoDB: Enable -Wconversion on clang and GCC The -Wconversion in GCC seems to be stricter than in clang. GCC at least since version 4.4.7 issues truncation warnings for assignments to bitfields, while clang 10 appears to only issue warnings when the sizes in bytes rounded to the nearest integer powers of 2 are different. Before GCC 10.0.0, -Wconversion required more casts and would not allow some operations, such as x<<=1 or x+=1 on a data type that is narrower than int. GCC 5 (but not GCC 4, GCC 6, or any later version) is complaining about x|=y even when x and y are compatible types that are narrower than int. Hence, we must rewrite some x|=y as x=static_cast<byte>(x|y) or similar, or we must disable -Wconversion. In GCC 6 and later, the warning for assigning wider to bitfields that are narrower than 8, 16, or 32 bits can be suppressed by applying a bitwise & with the exact bitmask of the bitfield. For older GCC, we must disable -Wconversion for GCC 4 or 5 in such cases. The bitwise negation operator appears to promote short integers to a wider type, and hence we must add explicit truncation casts around them. Microsoft Visual C does not allow a static_cast to truncate a constant, such as static_cast<byte>(1) truncating int. Hence, we will use the constructor-style cast byte(~1) for such cases. This has been tested at least with GCC 4.8.5, 5.4.0, 7.4.0, 9.2.1, 10.0.0, clang 9.0.1, 10.0.0, and MSVC 14.22.27905 (Microsoft Visual Studio 2019) on 64-bit and 32-bit targets (IA-32, AMD64, POWER 8, POWER 9, ARMv8).
2020-03-12 18:46:41 +01:00
uint16_t
dict_index_get_n_unique_in_tree_nonleaf(
const dict_index_t* index)
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Gets the number of user-defined ordering fields in the index. In the internal
representation we add the row id to the ordering fields to make all indexes
unique, but this function returns the number of fields the user defined
in the index as ordering fields.
@return number of fields */
UNIV_INLINE
MDEV-21907: InnoDB: Enable -Wconversion on clang and GCC The -Wconversion in GCC seems to be stricter than in clang. GCC at least since version 4.4.7 issues truncation warnings for assignments to bitfields, while clang 10 appears to only issue warnings when the sizes in bytes rounded to the nearest integer powers of 2 are different. Before GCC 10.0.0, -Wconversion required more casts and would not allow some operations, such as x<<=1 or x+=1 on a data type that is narrower than int. GCC 5 (but not GCC 4, GCC 6, or any later version) is complaining about x|=y even when x and y are compatible types that are narrower than int. Hence, we must rewrite some x|=y as x=static_cast<byte>(x|y) or similar, or we must disable -Wconversion. In GCC 6 and later, the warning for assigning wider to bitfields that are narrower than 8, 16, or 32 bits can be suppressed by applying a bitwise & with the exact bitmask of the bitfield. For older GCC, we must disable -Wconversion for GCC 4 or 5 in such cases. The bitwise negation operator appears to promote short integers to a wider type, and hence we must add explicit truncation casts around them. Microsoft Visual C does not allow a static_cast to truncate a constant, such as static_cast<byte>(1) truncating int. Hence, we will use the constructor-style cast byte(~1) for such cases. This has been tested at least with GCC 4.8.5, 5.4.0, 7.4.0, 9.2.1, 10.0.0, clang 9.0.1, 10.0.0, and MSVC 14.22.27905 (Microsoft Visual Studio 2019) on 64-bit and 32-bit targets (IA-32, AMD64, POWER 8, POWER 9, ARMv8).
2020-03-12 18:46:41 +01:00
uint16_t
dict_index_get_n_ordering_defined_by_user(
/*======================================*/
const dict_index_t* index) /*!< in: an internal representation
of index (in the dictionary cache) */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
#ifdef UNIV_DEBUG
/********************************************************************//**
Gets the nth field of an index.
@return pointer to field object */
UNIV_INLINE
dict_field_t*
dict_index_get_nth_field(
/*=====================*/
const dict_index_t* index, /*!< in: index */
ulint pos) /*!< in: position of field */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
#else /* UNIV_DEBUG */
# define dict_index_get_nth_field(index, pos) ((index)->fields + (pos))
#endif /* UNIV_DEBUG */
/********************************************************************//**
Gets pointer to the nth column in an index.
@return column */
UNIV_INLINE
const dict_col_t*
dict_index_get_nth_col(
/*===================*/
const dict_index_t* index, /*!< in: index */
ulint pos) /*!< in: position of the field */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Gets the column number of the nth field in an index.
@return column number */
UNIV_INLINE
ulint
dict_index_get_nth_col_no(
/*======================*/
const dict_index_t* index, /*!< in: index */
ulint pos) /*!< in: position of the field */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Looks for column n in an index.
@return position in internal representation of the index;
ULINT_UNDEFINED if not contained */
UNIV_INLINE
ulint
dict_index_get_nth_col_pos(
/*=======================*/
const dict_index_t* index, /*!< in: index */
ulint n, /*!< in: column number */
ulint* prefix_col_pos) /*!< out: col num if prefix */
MY_ATTRIBUTE((nonnull(1), warn_unused_result));
/** Looks for column n in an index.
@param[in] index index
@param[in] n column number
@param[in] inc_prefix true=consider column prefixes too
@param[in] is_virtual true==virtual column
@return position in internal representation of the index;
ULINT_UNDEFINED if not contained */
ulint
dict_index_get_nth_col_or_prefix_pos(
const dict_index_t* index, /*!< in: index */
ulint n, /*!< in: column number */
bool inc_prefix, /*!< in: TRUE=consider
column prefixes too */
bool is_virtual, /*!< in: is a virtual column
*/
ulint* prefix_col_pos) /*!< out: col num if prefix
*/
__attribute__((warn_unused_result));
/********************************************************************//**
Looks for a matching field in an index. The column has to be the same. The
column in index must be complete, or must contain a prefix longer than the
column in index2. That is, we must be able to construct the prefix in index2
from the prefix in index.
@return position in internal representation of the index;
ULINT_UNDEFINED if not contained */
ulint
dict_index_get_nth_field_pos(
/*=========================*/
const dict_index_t* index, /*!< in: index from which to search */
const dict_index_t* index2, /*!< in: index */
ulint n) /*!< in: field number in index2 */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Looks for column n position in the clustered index.
@return position in internal representation of the clustered index */
unsigned
dict_table_get_nth_col_pos(
/*=======================*/
const dict_table_t* table, /*!< in: table */
ulint n, /*!< in: column number */
ulint* prefix_col_pos) /*!< out: col num if prefix */
MY_ATTRIBUTE((nonnull(1), warn_unused_result));
/*******************************************************************//**
Adds a column to index. */
void
dict_index_add_col(
/*===============*/
dict_index_t* index, /*!< in/out: index */
const dict_table_t* table, /*!< in: table */
dict_col_t* col, /*!< in: column */
ulint prefix_len) /*!< in: column prefix length */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/*******************************************************************//**
Copies types of fields contained in index to tuple. */
void
dict_index_copy_types(
/*==================*/
dtuple_t* tuple, /*!< in/out: data tuple */
const dict_index_t* index, /*!< in: index */
ulint n_fields) /*!< in: number of
field types to copy */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/*********************************************************************//**
Gets the field column.
@return field->col, pointer to the table column */
UNIV_INLINE
const dict_col_t*
dict_field_get_col(
/*===============*/
const dict_field_t* field) /*!< in: index field */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/**********************************************************************//**
Returns an index object if it is found in the dictionary cache.
@return index, NULL if not found */
dict_index_t*
dict_index_get_if_in_cache_low(
/*===========================*/
index_id_t index_id) /*!< in: index id */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((warn_unused_result));
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
#ifdef UNIV_DEBUG
/**********************************************************************//**
Returns an index object if it is found in the dictionary cache.
@return index, NULL if not found */
dict_index_t*
dict_index_get_if_in_cache(
/*=======================*/
index_id_t index_id) /*!< in: index id */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((warn_unused_result));
/**********************************************************************//**
Checks that a tuple has n_fields_cmp value in a sensible range, so that
no comparison can occur with the page number field in a node pointer.
@return TRUE if ok */
ibool
dict_index_check_search_tuple(
/*==========================*/
const dict_index_t* index, /*!< in: index tree */
const dtuple_t* tuple) /*!< in: tuple used in a search */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Whether and when to allow temporary index names */
enum check_name {
/** Require all indexes to be complete. */
CHECK_ALL_COMPLETE,
/** Allow aborted online index creation. */
CHECK_ABORTED_OK,
/** Allow partial indexes to exist. */
CHECK_PARTIAL_OK
};
/**********************************************************************//**
Check for duplicate index entries in a table [using the index name] */
void
dict_table_check_for_dup_indexes(
/*=============================*/
const dict_table_t* table, /*!< in: Check for dup indexes
in this table */
enum check_name check) /*!< in: whether and when to allow
temporary index names */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
#endif /* UNIV_DEBUG */
/**********************************************************************//**
Builds a node pointer out of a physical record and a page number.
@return own: node pointer */
dtuple_t*
dict_index_build_node_ptr(
/*======================*/
const dict_index_t* index, /*!< in: index */
const rec_t* rec, /*!< in: record for which to build node
pointer */
ulint page_no,/*!< in: page number to put in node
pointer */
mem_heap_t* heap, /*!< in: memory heap where pointer
created */
ulint level) /*!< in: level of rec in tree:
0 means leaf level */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Convert a physical record into a search tuple.
@param[in] rec index record (not necessarily in an index page)
@param[in] index index
@param[in] leaf whether rec is in a leaf page
@param[in] n_fields number of data fields
@param[in,out] heap memory heap for allocation
@return own: data tuple */
dtuple_t*
MDEV-11369 Instant ADD COLUMN for InnoDB For InnoDB tables, adding, dropping and reordering columns has required a rebuild of the table and all its indexes. Since MySQL 5.6 (and MariaDB 10.0) this has been supported online (LOCK=NONE), allowing concurrent modification of the tables. This work revises the InnoDB ROW_FORMAT=REDUNDANT, ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC so that columns can be appended instantaneously, with only minor changes performed to the table structure. The counter innodb_instant_alter_column in INFORMATION_SCHEMA.GLOBAL_STATUS is incremented whenever a table rebuild operation is converted into an instant ADD COLUMN operation. ROW_FORMAT=COMPRESSED tables will not support instant ADD COLUMN. Some usability limitations will be addressed in subsequent work: MDEV-13134 Introduce ALTER TABLE attributes ALGORITHM=NOCOPY and ALGORITHM=INSTANT MDEV-14016 Allow instant ADD COLUMN, ADD INDEX, LOCK=NONE The format of the clustered index (PRIMARY KEY) is changed as follows: (1) The FIL_PAGE_TYPE of the root page will be FIL_PAGE_TYPE_INSTANT, and a new field PAGE_INSTANT will contain the original number of fields in the clustered index ('core' fields). If instant ADD COLUMN has not been used or the table becomes empty, or the very first instant ADD COLUMN operation is rolled back, the fields PAGE_INSTANT and FIL_PAGE_TYPE will be reset to 0 and FIL_PAGE_INDEX. (2) A special 'default row' record is inserted into the leftmost leaf, between the page infimum and the first user record. This record is distinguished by the REC_INFO_MIN_REC_FLAG, and it is otherwise in the same format as records that contain values for the instantly added columns. This 'default row' always has the same number of fields as the clustered index according to the table definition. The values of 'core' fields are to be ignored. For other fields, the 'default row' will contain the default values as they were during the ALTER TABLE statement. (If the column default values are changed later, those values will only be stored in the .frm file. The 'default row' will contain the original evaluated values, which must be the same for every row.) The 'default row' must be completely hidden from higher-level access routines. Assertions have been added to ensure that no 'default row' is ever present in the adaptive hash index or in locked records. The 'default row' is never delete-marked. (3) In clustered index leaf page records, the number of fields must reside between the number of 'core' fields (dict_index_t::n_core_fields introduced in this work) and dict_index_t::n_fields. If the number of fields is less than dict_index_t::n_fields, the missing fields are replaced with the column value of the 'default row'. Note: The number of fields in the record may shrink if some of the last instantly added columns are updated to the value that is in the 'default row'. The function btr_cur_trim() implements this 'compression' on update and rollback; dtuple::trim() implements it on insert. (4) In ROW_FORMAT=COMPACT and ROW_FORMAT=DYNAMIC records, the new status value REC_STATUS_COLUMNS_ADDED will indicate the presence of a new record header that will encode n_fields-n_core_fields-1 in 1 or 2 bytes. (In ROW_FORMAT=REDUNDANT records, the record header always explicitly encodes the number of fields.) We introduce the undo log record type TRX_UNDO_INSERT_DEFAULT for covering the insert of the 'default row' record when instant ADD COLUMN is used for the first time. Subsequent instant ADD COLUMN can use TRX_UNDO_UPD_EXIST_REC. This is joint work with Vin Chen (陈福荣) from Tencent. The design that was discussed in April 2017 would not have allowed import or export of data files, because instead of the 'default row' it would have introduced a data dictionary table. The test rpl.rpl_alter_instant is exactly as contributed in pull request #408. The test innodb.instant_alter is based on a contributed test. The redo log record format changes for ROW_FORMAT=DYNAMIC and ROW_FORMAT=COMPACT are as contributed. (With this change present, crash recovery from MariaDB 10.3.1 will fail in spectacular ways!) Also the semantics of higher-level redo log records that modify the PAGE_INSTANT field is changed. The redo log format version identifier was already changed to LOG_HEADER_FORMAT_CURRENT=103 in MariaDB 10.3.1. Everything else has been rewritten by me. Thanks to Elena Stepanova, the code has been tested extensively. When rolling back an instant ADD COLUMN operation, we must empty the PAGE_FREE list after deleting or shortening the 'default row' record, by calling either btr_page_empty() or btr_page_reorganize(). We must know the size of each entry in the PAGE_FREE list. If rollback left a freed copy of the 'default row' in the PAGE_FREE list, we would be unable to determine its size (if it is in ROW_FORMAT=COMPACT or ROW_FORMAT=DYNAMIC) because it would contain more fields than the rolled-back definition of the clustered index. UNIV_SQL_DEFAULT: A new special constant that designates an instantly added column that is not present in the clustered index record. len_is_stored(): Check if a length is an actual length. There are two magic length values: UNIV_SQL_DEFAULT, UNIV_SQL_NULL. dict_col_t::def_val: The 'default row' value of the column. If the column is not added instantly, def_val.len will be UNIV_SQL_DEFAULT. dict_col_t: Add the accessors is_virtual(), is_nullable(), is_instant(), instant_value(). dict_col_t::remove_instant(): Remove the 'instant ADD' status of a column. dict_col_t::name(const dict_table_t& table): Replaces dict_table_get_col_name(). dict_index_t::n_core_fields: The original number of fields. For secondary indexes and if instant ADD COLUMN has not been used, this will be equal to dict_index_t::n_fields. dict_index_t::n_core_null_bytes: Number of bytes needed to represent the null flags; usually equal to UT_BITS_IN_BYTES(n_nullable). dict_index_t::NO_CORE_NULL_BYTES: Magic value signalling that n_core_null_bytes was not initialized yet from the clustered index root page. dict_index_t: Add the accessors is_instant(), is_clust(), get_n_nullable(), instant_field_value(). dict_index_t::instant_add_field(): Adjust clustered index metadata for instant ADD COLUMN. dict_index_t::remove_instant(): Remove the 'instant ADD' status of a clustered index when the table becomes empty, or the very first instant ADD COLUMN operation is rolled back. dict_table_t: Add the accessors is_instant(), is_temporary(), supports_instant(). dict_table_t::instant_add_column(): Adjust metadata for instant ADD COLUMN. dict_table_t::rollback_instant(): Adjust metadata on the rollback of instant ADD COLUMN. prepare_inplace_alter_table_dict(): First create the ctx->new_table, and only then decide if the table really needs to be rebuilt. We must split the creation of table or index metadata from the creation of the dictionary table records and the creation of the data. In this way, we can transform a table-rebuilding operation into an instant ADD COLUMN operation. Dictionary objects will only be added to cache when table rebuilding or index creation is needed. The ctx->instant_table will never be added to cache. dict_table_t::add_to_cache(): Modified and renamed from dict_table_add_to_cache(). Do not modify the table metadata. Let the callers invoke dict_table_add_system_columns() and if needed, set can_be_evicted. dict_create_sys_tables_tuple(), dict_create_table_step(): Omit the system columns (which will now exist in the dict_table_t object already at this point). dict_create_table_step(): Expect the callers to invoke dict_table_add_system_columns(). pars_create_table(): Before creating the table creation execution graph, invoke dict_table_add_system_columns(). row_create_table_for_mysql(): Expect all callers to invoke dict_table_add_system_columns(). create_index_dict(): Replaces row_merge_create_index_graph(). innodb_update_n_cols(): Renamed from innobase_update_n_virtual(). Call my_error() if an error occurs. btr_cur_instant_init(), btr_cur_instant_init_low(), btr_cur_instant_root_init(): Load additional metadata from the clustered index and set dict_index_t::n_core_null_bytes. This is invoked when table metadata is first loaded into the data dictionary. dict_boot(): Initialize n_core_null_bytes for the four hard-coded dictionary tables. dict_create_index_step(): Initialize n_core_null_bytes. This is executed as part of CREATE TABLE. dict_index_build_internal_clust(): Initialize n_core_null_bytes to NO_CORE_NULL_BYTES if table->supports_instant(). row_create_index_for_mysql(): Initialize n_core_null_bytes for CREATE TEMPORARY TABLE. commit_cache_norebuild(): Call the code to rename or enlarge columns in the cache only if instant ADD COLUMN is not being used. (Instant ADD COLUMN would copy all column metadata from instant_table to old_table, including the names and lengths.) PAGE_INSTANT: A new 13-bit field for storing dict_index_t::n_core_fields. This is repurposing the 16-bit field PAGE_DIRECTION, of which only the least significant 3 bits were used. The original byte containing PAGE_DIRECTION will be accessible via the new constant PAGE_DIRECTION_B. page_get_instant(), page_set_instant(): Accessors for the PAGE_INSTANT. page_ptr_get_direction(), page_get_direction(), page_ptr_set_direction(): Accessors for PAGE_DIRECTION. page_direction_reset(): Reset PAGE_DIRECTION, PAGE_N_DIRECTION. page_direction_increment(): Increment PAGE_N_DIRECTION and set PAGE_DIRECTION. rec_get_offsets(): Use the 'leaf' parameter for non-debug purposes, and assume that heap_no is always set. Initialize all dict_index_t::n_fields for ROW_FORMAT=REDUNDANT records, even if the record contains fewer fields. rec_offs_make_valid(): Add the parameter 'leaf'. rec_copy_prefix_to_dtuple(): Assert that the tuple is only built on the core fields. Instant ADD COLUMN only applies to the clustered index, and we should never build a search key that has more than the PRIMARY KEY and possibly DB_TRX_ID,DB_ROLL_PTR. All these columns are always present. dict_index_build_data_tuple(): Remove assertions that would be duplicated in rec_copy_prefix_to_dtuple(). rec_init_offsets(): Support ROW_FORMAT=REDUNDANT records whose number of fields is between n_core_fields and n_fields. cmp_rec_rec_with_match(): Implement the comparison between two MIN_REC_FLAG records. trx_t::in_rollback: Make the field available in non-debug builds. trx_start_for_ddl_low(): Remove dangerous error-tolerance. A dictionary transaction must be flagged as such before it has generated any undo log records. This is because trx_undo_assign_undo() will mark the transaction as a dictionary transaction in the undo log header right before the very first undo log record is being written. btr_index_rec_validate(): Account for instant ADD COLUMN row_undo_ins_remove_clust_rec(): On the rollback of an insert into SYS_COLUMNS, revert instant ADD COLUMN in the cache by removing the last column from the table and the clustered index. row_search_on_row_ref(), row_undo_mod_parse_undo_rec(), row_undo_mod(), trx_undo_update_rec_get_update(): Handle the 'default row' as a special case. dtuple_t::trim(index): Omit a redundant suffix of an index tuple right before insert or update. After instant ADD COLUMN, if the last fields of a clustered index tuple match the 'default row', there is no need to store them. While trimming the entry, we must hold a page latch, so that the table cannot be emptied and the 'default row' be deleted. btr_cur_optimistic_update(), btr_cur_pessimistic_update(), row_upd_clust_rec_by_insert(), row_ins_clust_index_entry_low(): Invoke dtuple_t::trim() if needed. row_ins_clust_index_entry(): Restore dtuple_t::n_fields after calling row_ins_clust_index_entry_low(). rec_get_converted_size(), rec_get_converted_size_comp(): Allow the number of fields to be between n_core_fields and n_fields. Do not support infimum,supremum. They are never supposed to be stored in dtuple_t, because page creation nowadays uses a lower-level method for initializing them. rec_convert_dtuple_to_rec_comp(): Assign the status bits based on the number of fields. btr_cur_trim(): In an update, trim the index entry as needed. For the 'default row', handle rollback specially. For user records, omit fields that match the 'default row'. btr_cur_optimistic_delete_func(), btr_cur_pessimistic_delete(): Skip locking and adaptive hash index for the 'default row'. row_log_table_apply_convert_mrec(): Replace 'default row' values if needed. In the temporary file that is applied by row_log_table_apply(), we must identify whether the records contain the extra header for instantly added columns. For now, we will allocate an additional byte for this for ROW_T_INSERT and ROW_T_UPDATE records when the source table has been subject to instant ADD COLUMN. The ROW_T_DELETE records are fine, as they will be converted and will only contain 'core' columns (PRIMARY KEY and some system columns) that are converted from dtuple_t. rec_get_converted_size_temp(), rec_init_offsets_temp(), rec_convert_dtuple_to_temp(): Add the parameter 'status'. REC_INFO_DEFAULT_ROW = REC_INFO_MIN_REC_FLAG | REC_STATUS_COLUMNS_ADDED: An info_bits constant for distinguishing the 'default row' record. rec_comp_status_t: An enum of the status bit values. rec_leaf_format: An enum that replaces the bool parameter of rec_init_offsets_comp_ordinary().
2017-10-06 06:00:05 +02:00
dict_index_build_data_tuple(
const rec_t* rec,
const dict_index_t* index,
bool leaf,
ulint n_fields,
mem_heap_t* heap)
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Gets the page number of the root of the index tree.
@return page number */
UNIV_INLINE
uint32_t
dict_index_get_page(
/*================*/
const dict_index_t* tree) /*!< in: index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Returns free space reserved for future updates of records. This is
relevant only in the case of many consecutive inserts, as updates
which make the records bigger might fragment the index.
@return number of free bytes on page, reserved for updates */
UNIV_INLINE
ulint
dict_index_get_space_reserve(void);
/*==============================*/
/* Online index creation @{ */
/********************************************************************//**
Gets the status of online index creation.
@return the status */
UNIV_INLINE
enum online_index_status
dict_index_get_online_status(
/*=========================*/
const dict_index_t* index) /*!< in: secondary index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/********************************************************************//**
Sets the status of online index creation. */
UNIV_INLINE
void
dict_index_set_online_status(
/*=========================*/
dict_index_t* index, /*!< in/out: index */
enum online_index_status status) /*!< in: status */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/********************************************************************//**
Determines if a secondary index is being or has been created online,
or if the table is being rebuilt online, allowing concurrent modifications
to the table.
@retval true if the index is being or has been built online, or
if this is a clustered index and the table is being or has been rebuilt online
@retval false if the index has been created or the table has been
rebuilt completely */
UNIV_INLINE
bool
dict_index_is_online_ddl(
/*=====================*/
const dict_index_t* index) /*!< in: index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*********************************************************************//**
Calculates the minimum record length in an index. */
ulint
dict_index_calc_min_rec_len(
/*========================*/
const dict_index_t* index) /*!< in: index */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
2014-12-22 15:53:17 +01:00
/********************************************************************//**
Checks if the database name in two table names is the same.
@return TRUE if same db name */
ibool
dict_tables_have_same_db(
/*=====================*/
const char* name1, /*!< in: table name in the form
dbname '/' tablename */
const char* name2) /*!< in: table name in the form
dbname '/' tablename */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/** Get an index by name.
@param[in] table the table where to look for the index
@param[in] name the index name to look for
@return index, NULL if does not exist */
dict_index_t*
dict_table_get_index_on_name(dict_table_t* table, const char* name)
MY_ATTRIBUTE((warn_unused_result));
/** Get an index by name.
@param[in] table the table where to look for the index
@param[in] name the index name to look for
@return index, NULL if does not exist */
inline
const dict_index_t*
dict_table_get_index_on_name(const dict_table_t* table, const char* name)
{
return dict_table_get_index_on_name(const_cast<dict_table_t*>(table),
name);
}
/***************************************************************
Check whether a column exists in an FTS index. */
UNIV_INLINE
ulint
dict_table_is_fts_column(
/*=====================*/
/* out: ULINT_UNDEFINED if no match else
the offset within the vector */
ib_vector_t* indexes,/* in: vector containing only FTS indexes */
ulint col_no, /* in: col number to search for */
bool is_virtual)/*!< in: whether it is a virtual column */
MY_ATTRIBUTE((warn_unused_result));
/** Looks for an index with the given id given a table instance.
@param[in] table table instance
@param[in] id index id
@return index or NULL */
dict_index_t*
dict_table_find_index_on_id(
const dict_table_t* table,
index_id_t id)
MY_ATTRIBUTE((nonnull(1)));
/** Maximum number of columns in a foreign key constraint. Please Note MySQL
has a much lower limit on the number of columns allowed in a foreign key
constraint */
#define MAX_NUM_FK_COLUMNS 500
/* Buffers for storing detailed information about the latest foreign key
and unique key errors */
extern FILE* dict_foreign_err_file;
extern mysql_mutex_t dict_foreign_err_mutex;
/** InnoDB data dictionary cache */
class dict_sys_t
{
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** The my_hrtime_coarse().val of the oldest lock_wait() start, or 0 */
std::atomic<ulonglong> latch_ex_wait_start;
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
/** the rw-latch protecting the data dictionary cache */
alignas(CPU_LEVEL1_DCACHE_LINESIZE) srw_lock latch;
#ifdef UNIV_DEBUG
/** whether latch is being held in exclusive mode (by any thread) */
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
Atomic_relaxed<pthread_t> latch_ex;
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** number of S-latch holders */
Atomic_counter<uint32_t> latch_readers;
#endif
public:
/** Indexes of SYS_TABLE[] */
enum
{
SYS_TABLES= 0,
SYS_INDEXES,
SYS_COLUMNS,
SYS_FIELDS,
SYS_FOREIGN,
SYS_FOREIGN_COLS,
SYS_VIRTUAL
};
/** System table names */
static const span<const char> SYS_TABLE[];
/** all tables (persistent and temporary), hashed by name */
hash_table_t table_hash;
/** hash table of persistent table IDs */
hash_table_t table_id_hash;
/** the SYS_TABLES table */
dict_table_t *sys_tables;
/** the SYS_COLUMNS table */
dict_table_t *sys_columns;
/** the SYS_INDEXES table */
dict_table_t *sys_indexes;
/** the SYS_FIELDS table */
dict_table_t *sys_fields;
/** the SYS_FOREIGN table */
dict_table_t *sys_foreign;
/** the SYS_FOREIGN_COLS table */
dict_table_t *sys_foreign_cols;
/** the SYS_VIRTUAL table */
dict_table_t *sys_virtual;
/** @return whether all non-hard-coded system tables exist */
bool sys_tables_exist() const
{ return UNIV_LIKELY(sys_foreign && sys_foreign_cols && sys_virtual); }
/** list of persistent tables that can be evicted */
UT_LIST_BASE_NODE_T(dict_table_t) table_LRU;
/** list of persistent tables that cannot be evicted */
UT_LIST_BASE_NODE_T(dict_table_t) table_non_LRU;
private:
bool m_initialised= false;
/** the sequence of temporary table IDs */
std::atomic<table_id_t> temp_table_id{DICT_HDR_FIRST_ID};
/** hash table of temporary table IDs */
hash_table_t temp_id_hash;
/** the next value of DB_ROW_ID, backed by DICT_HDR_ROW_ID
(FIXME: remove this, and move to dict_table_t) */
Atomic_relaxed<row_id_t> row_id;
/** The synchronization interval of row_id */
static constexpr size_t ROW_ID_WRITE_MARGIN= 256;
public:
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** Diagnostic message for exceeding the lock_wait() timeout */
static const char fatal_msg[];
/** @return A new value for GEN_CLUST_INDEX(DB_ROW_ID) */
inline row_id_t get_new_row_id();
/** Ensure that row_id is not smaller than id, on IMPORT TABLESPACE */
inline void update_row_id(row_id_t id);
/** Recover the global DB_ROW_ID sequence on database startup */
void recover_row_id(row_id_t id)
{
row_id= ut_uint64_align_up(id, ROW_ID_WRITE_MARGIN) + ROW_ID_WRITE_MARGIN;
}
/** @return a new temporary table ID */
table_id_t acquire_temporary_table_id()
{
return temp_table_id.fetch_add(1, std::memory_order_relaxed);
}
MDEV-17794 Do not assign persistent ID for temporary tables InnoDB in MySQL 5.7 introduced two new parameters to the function dict_hdr_get_new_id(), to allow redo logging to be disabled when assigning identifiers to temporary tables or during the backup-unfriendly TRUNCATE TABLE that was replaced in MariaDB by MDEV-13564. Now that MariaDB 10.4.0 removed the crash recovery code for the backup-unfriendly TRUNCATE, we can revert dict_hdr_get_new_id() to be used only for persistent data structures. dict_table_assign_new_id(): Remove. This was a simple 2-line function that was called from few places. dict_table_open_on_id_low(): Declare in the only file where it is called. dict_sys_t::temp_id_hash: A separate lookup table for temporary tables. Table names will be in the common dict_sys_t::table_hash. dict_sys_t::get_temporary_table_id(): Assign a temporary table ID. dict_sys_t::get_table(): Look up a persistent table. dict_sys_t::get_temporary_table(): Look up a temporary table. dict_sys_t::temp_table_id: The sequence of temporary table identifiers. Starts from DICT_HDR_FIRST_ID, so that we can continue to simply compare dict_table_t::id to a few constants for the persistent hard-coded data dictionary tables. undo_node_t::state: Distinguish temporary and persistent tables. lock_check_dict_lock(), lock_get_table_id(): Assert that there cannot be locks on temporary tables. row_rec_to_index_entry_impl(): Assert that there cannot be metadata records on temporary tables. row_undo_ins_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_ins(): Add some assertions. row_undo_mod_clust(), row_undo_mod(): Do not assign node->state. Let row_undo() do that. row_undo_mod_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_try_truncate(): Renamed and simplified from trx_roll_try_truncate(). row_undo_rec_get(): Replaces trx_roll_pop_top_rec_of_trx() and trx_roll_pop_top_rec(). Fetch an undo log record, and assign undo->state accordingly. trx_undo_truncate_end(): Acquire the rseg->mutex only for the minimum required duration, and release it between mini-transactions.
2018-11-22 14:36:50 +01:00
/** Look up a temporary table.
@param id temporary table ID
@return temporary table
@retval nullptr if the table does not exist
(should only happen during the rollback of CREATE...SELECT) */
dict_table_t *acquire_temporary_table(table_id_t id)
{
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_ad(frozen());
dict_table_t *table;
ulint fold = ut_fold_ull(id);
HASH_SEARCH(id_hash, &temp_id_hash, fold, dict_table_t*, table,
ut_ad(table->cached), table->id == id);
if (UNIV_LIKELY(table != nullptr))
{
DBUG_ASSERT(table->is_temporary());
DBUG_ASSERT(table->id >= DICT_HDR_FIRST_ID);
table->acquire();
}
return table;
}
MDEV-17794 Do not assign persistent ID for temporary tables InnoDB in MySQL 5.7 introduced two new parameters to the function dict_hdr_get_new_id(), to allow redo logging to be disabled when assigning identifiers to temporary tables or during the backup-unfriendly TRUNCATE TABLE that was replaced in MariaDB by MDEV-13564. Now that MariaDB 10.4.0 removed the crash recovery code for the backup-unfriendly TRUNCATE, we can revert dict_hdr_get_new_id() to be used only for persistent data structures. dict_table_assign_new_id(): Remove. This was a simple 2-line function that was called from few places. dict_table_open_on_id_low(): Declare in the only file where it is called. dict_sys_t::temp_id_hash: A separate lookup table for temporary tables. Table names will be in the common dict_sys_t::table_hash. dict_sys_t::get_temporary_table_id(): Assign a temporary table ID. dict_sys_t::get_table(): Look up a persistent table. dict_sys_t::get_temporary_table(): Look up a temporary table. dict_sys_t::temp_table_id: The sequence of temporary table identifiers. Starts from DICT_HDR_FIRST_ID, so that we can continue to simply compare dict_table_t::id to a few constants for the persistent hard-coded data dictionary tables. undo_node_t::state: Distinguish temporary and persistent tables. lock_check_dict_lock(), lock_get_table_id(): Assert that there cannot be locks on temporary tables. row_rec_to_index_entry_impl(): Assert that there cannot be metadata records on temporary tables. row_undo_ins_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_ins(): Add some assertions. row_undo_mod_clust(), row_undo_mod(): Do not assign node->state. Let row_undo() do that. row_undo_mod_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_try_truncate(): Renamed and simplified from trx_roll_try_truncate(). row_undo_rec_get(): Replaces trx_roll_pop_top_rec_of_trx() and trx_roll_pop_top_rec(). Fetch an undo log record, and assign undo->state accordingly. trx_undo_truncate_end(): Acquire the rseg->mutex only for the minimum required duration, and release it between mini-transactions.
2018-11-22 14:36:50 +01:00
/** Look up a persistent table.
@param id table ID
@return table
@retval nullptr if not cached */
dict_table_t *find_table(table_id_t id)
{
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_ad(frozen());
dict_table_t *table;
ulint fold= ut_fold_ull(id);
HASH_SEARCH(id_hash, &table_id_hash, fold, dict_table_t*, table,
ut_ad(table->cached), table->id == id);
DBUG_ASSERT(!table || !table->is_temporary());
return table;
}
MDEV-17794 Do not assign persistent ID for temporary tables InnoDB in MySQL 5.7 introduced two new parameters to the function dict_hdr_get_new_id(), to allow redo logging to be disabled when assigning identifiers to temporary tables or during the backup-unfriendly TRUNCATE TABLE that was replaced in MariaDB by MDEV-13564. Now that MariaDB 10.4.0 removed the crash recovery code for the backup-unfriendly TRUNCATE, we can revert dict_hdr_get_new_id() to be used only for persistent data structures. dict_table_assign_new_id(): Remove. This was a simple 2-line function that was called from few places. dict_table_open_on_id_low(): Declare in the only file where it is called. dict_sys_t::temp_id_hash: A separate lookup table for temporary tables. Table names will be in the common dict_sys_t::table_hash. dict_sys_t::get_temporary_table_id(): Assign a temporary table ID. dict_sys_t::get_table(): Look up a persistent table. dict_sys_t::get_temporary_table(): Look up a temporary table. dict_sys_t::temp_table_id: The sequence of temporary table identifiers. Starts from DICT_HDR_FIRST_ID, so that we can continue to simply compare dict_table_t::id to a few constants for the persistent hard-coded data dictionary tables. undo_node_t::state: Distinguish temporary and persistent tables. lock_check_dict_lock(), lock_get_table_id(): Assert that there cannot be locks on temporary tables. row_rec_to_index_entry_impl(): Assert that there cannot be metadata records on temporary tables. row_undo_ins_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_ins(): Add some assertions. row_undo_mod_clust(), row_undo_mod(): Do not assign node->state. Let row_undo() do that. row_undo_mod_parse_undo_rec(): Distinguish temporary and persistent tables. Move some assertions from the only caller. Return whether the table was found. row_undo_try_truncate(): Renamed and simplified from trx_roll_try_truncate(). row_undo_rec_get(): Replaces trx_roll_pop_top_rec_of_trx() and trx_roll_pop_top_rec(). Fetch an undo log record, and assign undo->state accordingly. trx_undo_truncate_end(): Acquire the rseg->mutex only for the minimum required duration, and release it between mini-transactions.
2018-11-22 14:36:50 +01:00
bool is_initialised() const { return m_initialised; }
/** Initialise the data dictionary cache. */
void create();
/** Close the data dictionary cache on shutdown. */
void close();
/** Resize the hash tables based on the current buffer pool size. */
void resize();
/** Add a table definition to the data dictionary cache */
inline void add(dict_table_t* table);
/** Remove a table definition from the data dictionary cache.
@param[in,out] table cached table definition to be evicted
@param[in] lru whether this is part of least-recently-used evictiono
@param[in] keep whether to keep (not free) the object */
void remove(dict_table_t* table, bool lru = false, bool keep = false);
#ifdef UNIV_DEBUG
/** Find a table */
template <bool in_lru> bool find(const dict_table_t *table)
{
ut_ad(table);
ut_ad(table->can_be_evicted == in_lru);
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_ad(frozen());
for (const dict_table_t* t= in_lru ? table_LRU.start : table_non_LRU.start;
t; t = UT_LIST_GET_NEXT(table_LRU, t))
{
if (t == table) return true;
ut_ad(t->can_be_evicted == in_lru);
}
return false;
}
/** Find a table */
bool find(const dict_table_t *table)
{
return table->can_be_evicted ? find<true>(table) : find<false>(table);
}
#endif
/** Move a table to the non-LRU list from the LRU list. */
void prevent_eviction(dict_table_t *table)
{
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_d(locked());
ut_ad(find(table));
if (!table->can_be_evicted)
return;
table->can_be_evicted= false;
UT_LIST_REMOVE(table_LRU, table);
UT_LIST_ADD_LAST(table_non_LRU, table);
}
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
#ifdef UNIV_DEBUG
/** @return whether any thread (not necessarily the current thread)
is holding the latch; that is, this check may return false
positives */
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
bool frozen() const { return latch_readers || latch_ex; }
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** @return whether any thread (not necessarily the current thread)
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
is holding a shared latch */
bool frozen_not_locked() const { return latch_readers; }
/** @return whether the current thread holds the exclusive latch */
bool locked() const { return latch_ex == pthread_self(); }
#endif
private:
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** Acquire the exclusive latch */
ATTRIBUTE_NOINLINE
void lock_wait(SRW_LOCK_ARGS(const char *file, unsigned line));
public:
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** @return the my_hrtime_coarse().val of the oldest lock_wait() start,
assuming that requests are served on a FIFO basis */
ulonglong oldest_wait() const
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
{ return latch_ex_wait_start.load(std::memory_order_relaxed); }
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** Exclusively lock the dictionary cache. */
void lock(SRW_LOCK_ARGS(const char *file, unsigned line))
{
if (latch.wr_lock_try())
{
ut_ad(!latch_readers);
ut_ad(!latch_ex);
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
ut_d(latch_ex= pthread_self());
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
}
else
lock_wait(SRW_LOCK_ARGS(file, line));
}
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
#ifdef UNIV_PFS_RWLOCK
/** Unlock the data dictionary cache. */
ATTRIBUTE_NOINLINE void unlock();
/** Acquire a shared lock on the dictionary cache. */
ATTRIBUTE_NOINLINE void freeze(const char *file, unsigned line);
/** Release a shared lock on the dictionary cache. */
ATTRIBUTE_NOINLINE void unfreeze();
#else
/** Unlock the data dictionary cache. */
void unlock()
{
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
ut_ad(latch_ex == pthread_self());
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_ad(!latch_readers);
MDEV-29504/MDEV-29849 TRUNCATE breaks FOREIGN KEY locking ha_innobase::referenced_by_foreign_key(): Protect the check with dict_sys.freeze(), to prevent races with TRUNCATE TABLE. The test innodb.instant_alter_crash has been adjusted for this additional locking. dict_table_is_referenced_by_foreign_key(): Removed (merged to the only caller). create_table_info_t::create_table(): Ignore missing indexes for FOREIGN KEY constraints if foreign_key_checks=0. create_table_info_t::create_table_update_dict(): Rewritten as a static function. Do not return any error. ha_innobase::create(): When trx!=nullptr and we are operating on a persistent table, do not rollback, commit, or release the data dictionary latch. ha_innobase::truncate(): Protect the entire critical section with an exclusive dict_sys.latch, so that ha_innobase::referenced_by_foreign_key() on referenced tables will return a consistent result. In case of a failure, invoke dict_load_foreigns() to restore also any FOREIGN KEY constraints. ha_innobase::free_foreign_key_create_info(): Define inline. lock_release(): Disregard innodb_evict_tables_on_commit_debug=ON when dict_sys.locked() holds. It would hold when fts_load_stopword() is invoked by create_table_info_t::create_table_update_dict(). dict_sys_t::locked(): Return whether the current thread is holding the exclusive dict_sys.latch. dict_sys_t::frozen_not_locked(): Return whether any thread is holding a shared dict_sys.latch. In the test main.mysql_upgrade, the InnoDB persistent statistics will no longer be recalculated in ha_innobase::open() as part of CHECK TABLE ... FOR UPGRADE. They were deleted earlier in the test. Tested by: Matthias Leich
2022-11-08 16:34:34 +01:00
ut_d(latch_ex= 0);
latch.wr_unlock();
}
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
/** Acquire a shared lock on the dictionary cache. */
void freeze()
{
latch.rd_lock();
ut_ad(!latch_ex);
ut_d(latch_readers++);
}
/** Release a shared lock on the dictionary cache. */
void unfreeze()
{
ut_ad(!latch_ex);
ut_ad(latch_readers--);
latch.rd_unlock();
}
#endif
/** Estimate the used memory occupied by the data dictionary
table and index objects.
@return number of bytes occupied */
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
TPOOL_SUPPRESS_TSAN ulint rough_size() const
{
MDEV-25919: Lock tables before acquiring dict_sys.latch In commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 (MDEV-25506 part 3) we introduced a "fake instant timeout" when a transaction would wait for a table or record lock while holding dict_sys.latch. This prevented a deadlock of the server but could cause bogus errors for operations on the InnoDB persistent statistics tables. A better fix is to ensure that whenever a transaction is being executed in the InnoDB internal SQL parser (which will for now require dict_sys.latch to be held), it will already have acquired all locks that could be required for the execution. So, we will acquire the following locks upfront, before acquiring dict_sys.latch: (1) MDL on the affected user table (acquired by the SQL layer) (2) If applicable (not for RENAME TABLE): InnoDB table lock (3) If persistent statistics are going to be modified: (3.a) MDL_SHARED on mysql.innodb_table_stats, mysql.innodb_index_stats (3.b) exclusive table locks on the statistics tables (4) Exclusive table locks on the InnoDB data dictionary tables (not needed in ANALYZE TABLE and the like) Note: Acquiring exclusive locks on the statistics tables may cause more locking conflicts between concurrent DDL operations. Notably, RENAME TABLE will lock the statistics tables even if no persistent statistics are enabled for the table. DROP DATABASE will only acquire locks on statistics tables if persistent statistics are enabled for the tables on which the SQL layer is invoking ha_innobase::delete_table(). For any "garbage collection" in innodb_drop_database(), a timeout while acquiring locks on the statistics tables will result in any statistics not being deleted for any tables that the SQL layer did not know about. If innodb_defragment=ON, information may be written to the statistics tables even for tables for which InnoDB persistent statistics are disabled. But, DROP TABLE will no longer attempt to delete that information if persistent statistics are not enabled for the table. This change should also fix the hangs related to InnoDB persistent statistics and STATS_AUTO_RECALC (MDEV-15020) as well as a bug that running ALTER TABLE on the statistics tables concurrently with running ALTER TABLE on InnoDB tables could cause trouble. lock_rec_enqueue_waiting(), lock_table_enqueue_waiting(): Do not issue a fake instant timeout error when the transaction is holding dict_sys.latch. Instead, assert that the dict_sys.latch is never being held here. lock_sys_tables(): A new function to acquire exclusive locks on all dictionary tables, in case DROP TABLE or similar operation is being executed. Locking non-hard-coded tables is optional to avoid a crash in row_merge_drop_temp_indexes(). The SYS_VIRTUAL table was introduced in MySQL 5.7 and MariaDB Server 10.2. Normally, we require all these dictionary tables to exist before executing any DDL, but the function row_merge_drop_temp_indexes() is an exception. When upgrading from MariaDB Server 10.1 or MySQL 5.6 or earlier, the table SYS_VIRTUAL would not exist at this point. ha_innobase::commit_inplace_alter_table(): Invoke log_write_up_to() while not holding dict_sys.latch. dict_sys_t::remove(), dict_table_close(): No longer try to drop index stubs that were left behind by aborted online ADD INDEX. Such indexes should be dropped from the InnoDB data dictionary by row_merge_drop_indexes() as part of the failed DDL operation. Stubs for aborted indexes may only be left behind in the data dictionary cache. dict_stats_fetch_from_ps(): Use a normal read-only transaction. ha_innobase::delete_table(), ha_innobase::truncate(), fts_lock_table(): While waiting for purge to stop using the table, do not hold dict_sys.latch. ha_innobase::delete_table(): Implement a work-around for the rollback of ALTER TABLE...ADD PARTITION. MDL_EXCLUSIVE would not be held if ALTER TABLE hits lock_wait_timeout while trying to upgrade the MDL due to a conflicting LOCK TABLES, such as in the first ALTER TABLE in the test case of Bug#53676 in parts.partition_special_innodb. Therefore, we must explicitly stop purge, because it would not be stopped by MDL. dict_stats_func(), btr_defragment_chunk(): Allocate a THD so that we can acquire MDL on the InnoDB persistent statistics tables. mysqltest_embedded: Invoke ha_pre_shutdown() before free_used_memory() in order to avoid ASAN heap-use-after-free related to acquire_thd(). trx_t::dict_operation_lock_mode: Changed the type to bool. row_mysql_lock_data_dictionary(), row_mysql_unlock_data_dictionary(): Implemented as macros. rollback_inplace_alter_table(): Apply an infinite timeout to lock waits. innodb_thd_increment_pending_ops(): Wrapper for thd_increment_pending_ops(). Never attempt async operation for InnoDB background threads, such as the trx_t::commit() in dict_stats_process_entry_from_recalc_pool(). lock_sys_t::cancel(trx_t*): Make dictionary transactions immune to KILL. lock_wait(): Make dictionary transactions immune to KILL, and to lock wait timeout when waiting for locks on dictionary tables. parts.partition_special_innodb: Use lock_wait_timeout=0 to instantly get ER_LOCK_WAIT_TIMEOUT. main.mdl: Filter out MDL on InnoDB persistent statistics tables Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:44 +02:00
/* No latch; this is a very crude approximation anyway */
ulint size = UT_LIST_GET_LEN(table_LRU) + UT_LIST_GET_LEN(table_non_LRU);
size *= sizeof(dict_table_t)
+ sizeof(dict_index_t) * 2
+ (sizeof(dict_col_t) + sizeof(dict_field_t)) * 10
+ sizeof(dict_field_t) * 5 /* total number of key fields */
+ 200; /* arbitrary, covering names and overhead */
size += (table_hash.n_cells + table_id_hash.n_cells +
temp_id_hash.n_cells) * sizeof(hash_cell_t);
return size;
}
/** Evict unused, unlocked tables from table_LRU.
@param half whether to consider half the tables only (instead of all)
@return number of tables evicted */
ulint evict_table_LRU(bool half);
/** Look up a table in the dictionary cache.
@param name table name
@return table handle
@retval nullptr if not found */
dict_table_t *find_table(const span<const char> &name) const
{
MDEV-24258 Merge dict_sys.mutex into dict_sys.latch In the parent commit, dict_sys.latch could theoretically have been replaced with a mutex. But, we can do better and merge dict_sys.mutex into dict_sys.latch. Generally, every occurrence of dict_sys.mutex_lock() will be replaced with dict_sys.lock(). The PERFORMANCE_SCHEMA instrumentation for dict_sys_mutex will be removed along with dict_sys.mutex. The dict_sys.latch will remain instrumented as dict_operation_lock. Some use of dict_sys.lock() will be replaced with dict_sys.freeze(), which we will reintroduce for the new shared mode. Most notably, concurrent table lookups are possible as long as the tables are present in the dict_sys cache. In particular, this will allow more concurrency among InnoDB purge workers. Because dict_sys.mutex will no longer 'throttle' the threads that purge InnoDB transaction history, a performance degradation may be observed unless innodb_purge_threads=1. The table cache eviction policy will become FIFO-like, similar to what happened to fil_system.LRU in commit 45ed9dd957eebc7fc84feb2509f4aa6baa908a95. The name of the list dict_sys.table_LRU will become somewhat misleading; that list contains tables that may be evicted, even though the eviction policy no longer is least-recently-used but first-in-first-out. (Note: Tables can never be evicted as long as locks exist on them or the tables are in use by some thread.) As demonstrated by the test perfschema.sxlock_func, there will be less contention on dict_sys.latch, because some previous use of exclusive latches will be replaced with shared latches. fts_parse_sql_no_dict_lock(): Replaced with pars_sql(). fts_get_table_name_prefix(): Merged to fts_optimize_create(). dict_stats_update_transient_for_index(): Deduplicated some code. ha_innobase::info_low(), dict_stats_stop_bg(): Use a combination of dict_sys.latch and table->stats_mutex_lock() to cover the changes of BG_STAT_SHOULD_QUIT, because the flag is being read in dict_stats_update_persistent() while not holding dict_sys.latch. row_discard_tablespace_for_mysql(): Protect stats_bg_flag by exclusive dict_sys.latch, like most other code does. row_quiesce_table_has_fts_index(): Remove unnecessary mutex acquisition. FLUSH TABLES...FOR EXPORT is protected by MDL. row_import::set_root_by_heuristic(): Remove unnecessary mutex acquisition. ALTER TABLE...IMPORT TABLESPACE is protected by MDL. row_ins_sec_index_entry_low(): Replace a call to dict_set_corrupted_index_cache_only(). Reads of index->type were not really protected by dict_sys.mutex, and writes (flagging an index corrupted) should be extremely rare. dict_stats_process_entry_from_defrag_pool(): Only freeze the dictionary, do not lock it exclusively. dict_stats_wait_bg_to_stop_using_table(), DICT_BG_YIELD: Remove trx. We can simply invoke dict_sys.unlock() and dict_sys.lock() directly. dict_acquire_mdl_shared()<trylock=false>: Assert that dict_sys.latch is only held in shared more, not exclusive mode. Only acquire it in exclusive mode if the table needs to be loaded to the cache. dict_sys_t::acquire(): Remove. Relocating elements in dict_sys.table_LRU would require holding an exclusive latch, which we want to avoid for performance reasons. dict_sys_t::allow_eviction(): Add the table first to dict_sys.table_LRU, to compensate for the removal of dict_sys_t::acquire(). This function is only invoked by INFORMATION_SCHEMA.INNODB_SYS_TABLESTATS. dict_table_open_on_id(), dict_table_open_on_name(): If dict_locked=false, try to acquire dict_sys.latch in shared mode. Only acquire the latch in exclusive mode if the table is not found in the cache. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:51:35 +02:00
ut_ad(frozen());
for (dict_table_t *table= static_cast<dict_table_t*>
(HASH_GET_FIRST(&table_hash, table_hash.calc_hash
(my_crc32c(0, name.data(), name.size()))));
table; table= table->name_hash)
if (strlen(table->name.m_name) == name.size() &&
!memcmp(table->name.m_name, name.data(), name.size()))
return table;
return nullptr;
}
/** Look up or load a table definition
@param name table name
@param ignore errors to ignore when loading the table definition
@return table handle
@retval nullptr if not found */
dict_table_t *load_table(const span<const char> &name,
dict_err_ignore_t ignore= DICT_ERR_IGNORE_NONE);
/** Attempt to load the system tables on startup
@return whether any discrepancy with the expected definition was found */
bool load_sys_tables();
/** Create or check system tables on startup */
dberr_t create_or_check_sys_tables();
};
/** the data dictionary cache */
extern dict_sys_t dict_sys;
/*********************************************************************//**
Converts a database and table name from filesystem encoding
(e.g. d@i1b/a@q1b@1Kc, same format as used in dict_table_t::name) in two
strings in UTF8 encoding (e.g. dцb and aюbØc). The output buffers must be
at least MAX_DB_UTF8_LEN and MAX_TABLE_UTF8_LEN bytes. */
void
dict_fs2utf8(
/*=========*/
const char* db_and_table, /*!< in: database and table names,
e.g. d@i1b/a@q1b@1Kc */
char* db_utf8, /*!< out: database name, e.g. dцb */
size_t db_utf8_size, /*!< in: dbname_utf8 size */
char* table_utf8, /*!< out: table name, e.g. aюbØc */
size_t table_utf8_size)/*!< in: table_utf8 size */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
MDEV-13542: Crashing on corrupted page is unhelpful The approach to handling corruption that was chosen by Oracle in commit 177d8b0c125b841c0650d27d735e3b87509dc286 is not really useful. Not only did it actually fail to prevent InnoDB from crashing, but it is making things worse by blocking attempts to rescue data from or rebuild a partially readable table. We will try to prevent crashes in a different way: by propagating errors up the call stack. We will never mark the clustered index persistently corrupted, so that data recovery may be attempted by reading from the table, or by rebuilding the table. This should also fix MDEV-13680 (crash on btr_page_alloc() failure); it was extensively tested with innodb_file_per_table=0 and a non-autoextend system tablespace. We should now avoid crashes in many cases, such as when a page cannot be read or allocated, or an inconsistency is detected when attempting to update multiple pages. We will not crash on double-free, such as on the recovery of DDL in system tablespace in case something was corrupted. Crashes on corrupted data are still possible. The fault injection mechanism that is introduced in the subsequent commit may help catch more of them. buf_page_import_corrupt_failure: Remove the fault injection, and instead corrupt some pages using Perl code in the tests. btr_cur_pessimistic_insert(): Always reserve extents (except for the change buffer), in order to prevent a subsequent allocation failure. btr_pcur_open_at_rnd_pos(): Merged to the only caller ibuf_merge_pages(). btr_assert_not_corrupted(), btr_corruption_report(): Remove. Similar checks are already part of btr_block_get(). FSEG_MAGIC_N_BYTES: Replaces FSEG_MAGIC_N_VALUE. dict_hdr_get(), trx_rsegf_get_new(), trx_undo_page_get(), trx_undo_page_get_s_latched(): Replaced with error-checking calls. trx_rseg_t::get(mtr_t*): Replaces trx_rsegf_get(). trx_rseg_header_create(): Let the caller update the TRX_SYS page if needed. trx_sys_create_sys_pages(): Merged with trx_sysf_create(). dict_check_tablespaces_and_store_max_id(): Do not access DICT_HDR_MAX_SPACE_ID, because it was already recovered in dict_boot(). Merge dict_check_sys_tables() with this function. dir_pathname(): Replaces os_file_make_new_pathname(). row_undo_ins_remove_sec(): Do not modify the undo page by adding a terminating NUL byte to the record. btr_decryption_failed(): Report decryption failures dict_set_corrupted_by_space(), dict_set_encrypted_by_space(), dict_set_corrupted_index_cache_only(): Remove. dict_set_corrupted(): Remove the constant parameter dict_locked=false. Never flag the clustered index corrupted in SYS_INDEXES, because that would deny further access to the table. It might be possible to repair the table by executing ALTER TABLE or OPTIMIZE TABLE, in case no B-tree leaf page is corrupted. dict_table_skip_corrupt_index(), dict_table_next_uncorrupted_index(), row_purge_skip_uncommitted_virtual_index(): Remove, and refactor the callers to read dict_index_t::type only once. dict_table_is_corrupted(): Remove. dict_index_t::is_btree(): Determine if the index is a valid B-tree. BUF_GET_NO_LATCH, BUF_EVICT_IF_IN_POOL: Remove. UNIV_BTR_DEBUG: Remove. Any inconsistency will no longer trigger assertion failures, but error codes being returned. buf_corrupt_page_release(): Replaced with a direct call to buf_pool.corrupted_evict(). fil_invalid_page_access_msg(): Never crash on an invalid read; let the caller of buf_page_get_gen() decide. btr_pcur_t::restore_position(): Propagate failure status to the caller by returning CORRUPTED. opt_search_plan_for_table(): Simplify the code. row_purge_del_mark(), row_purge_upd_exist_or_extern_func(), row_undo_ins_remove_sec_rec(), row_undo_mod_upd_del_sec(), row_undo_mod_del_mark_sec(): Avoid mem_heap_create()/mem_heap_free() when no secondary indexes exist. row_undo_mod_upd_exist_sec(): Simplify the code. row_upd_clust_step(), dict_load_table_one(): Return DB_TABLE_CORRUPT if the clustered index (and therefore the table) is corrupted, similar to what we do in row_insert_for_mysql(). fut_get_ptr(): Replace with buf_page_get_gen() calls. buf_page_get_gen(): Return nullptr and *err=DB_CORRUPTION if the page is marked as freed. For other modes than BUF_GET_POSSIBLY_FREED or BUF_PEEK_IF_IN_POOL this will trigger a debug assertion failure. For BUF_GET_POSSIBLY_FREED, we will return nullptr for freed pages, so that the callers can be simplified. The purge of transaction history will be a new user of BUF_GET_POSSIBLY_FREED, to avoid crashes on corrupted data. buf_page_get_low(): Never crash on a corrupted page, but simply return nullptr. fseg_page_is_allocated(): Replaces fseg_page_is_free(). fts_drop_common_tables(): Return an error if the transaction was rolled back. fil_space_t::set_corrupted(): Report a tablespace as corrupted if it was not reported already. fil_space_t::io(): Invoke fil_space_t::set_corrupted() to report out-of-bounds page access or other errors. Clean up mtr_t::page_lock() buf_page_get_low(): Validate the page identifier (to check for recently read corrupted pages) after acquiring the page latch. buf_page_t::read_complete(): Flag uninitialized (all-zero) pages with DB_FAIL. Return DB_PAGE_CORRUPTED on page number mismatch. mtr_t::defer_drop_ahi(): Renamed from mtr_defer_drop_ahi(). recv_sys_t::free_corrupted_page(): Only set_corrupt_fs() if any log records exist for the page. We do not mind if read-ahead produces corrupted (or all-zero) pages that were not actually needed during recovery. recv_recover_page(): Return whether the operation succeeded. recv_sys_t::recover_low(): Simplify the logic. Check for recovery error. Thanks to Matthias Leich for testing this extensively and to the authors of https://rr-project.org for making it easy to diagnose and fix any failures that were found during the testing.
2022-06-06 13:03:22 +02:00
/** Flag an index corrupted both in the data dictionary cache
MDEV-25919 preparation: Various cleanup que_eval_sql(): Remove the parameter lock_dict. The only caller with lock_dict=true was dict_stats_exec_sql(), which will now explicitly invoke dict_sys.lock() and dict_sys.unlock() by itself. row_import_cleanup(): Do not unnecessarily lock the dictionary. Concurrent access to the table during ALTER TABLE...IMPORT TABLESPACE is prevented by MDL and the fact that there cannot exist any undo log or change buffer records that would refer to the table or tablespace. row_import_for_mysql(): Do not unnecessarily lock the dictionary while accessing fil_system. Thanks to MDL_EXCLUSIVE that was acquired by the SQL layer, only one IMPORT may be in effect for the table name. row_quiesce_set_state(): Do not unnecessarily lock the dictionary. The dict_table_t::quiesce state is documented to be protected by all index latches, which we are acquiring. dict_table_close(): Introduce a simpler variant with fewer parameters. dict_table_close(): Reduce the amount of calls. We can simply invoke dict_table_t::release() on startup or in DDL operations, or when the table is inaccessible. In none of these cases, there is no need to invalidate the InnoDB persistent statistics. pars_info_t::graph_owns_us: Remove (unused). pars_info_free(): Define inline. fts_delete(), trx_t::evict_table(), row_prebuilt_free(), row_rename_table_for_mysql(): Simplify. row_mysql_lock_data_dictionary(): Remove some references; use dict_sys.lock() and dict_sys.unlock() instead. row_mysql_lock_table(): Remove. Use lock_table_for_trx() instead. ha_innobase::check_if_supported_inplace_alter(), row_create_table_for_mysql(): Simply assert dict_sys.sys_tables_exist(). In commit 49e2c8f0a6fefdeac50925f758090d6bd099768d and commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 srv_start() actually guarantees that the system tables will exist, or the server is in read-only mode, or startup will fail. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:20 +02:00
and in the system table SYS_INDEXES.
@param index index to be flagged as corrupted
MDEV-13542: Crashing on corrupted page is unhelpful The approach to handling corruption that was chosen by Oracle in commit 177d8b0c125b841c0650d27d735e3b87509dc286 is not really useful. Not only did it actually fail to prevent InnoDB from crashing, but it is making things worse by blocking attempts to rescue data from or rebuild a partially readable table. We will try to prevent crashes in a different way: by propagating errors up the call stack. We will never mark the clustered index persistently corrupted, so that data recovery may be attempted by reading from the table, or by rebuilding the table. This should also fix MDEV-13680 (crash on btr_page_alloc() failure); it was extensively tested with innodb_file_per_table=0 and a non-autoextend system tablespace. We should now avoid crashes in many cases, such as when a page cannot be read or allocated, or an inconsistency is detected when attempting to update multiple pages. We will not crash on double-free, such as on the recovery of DDL in system tablespace in case something was corrupted. Crashes on corrupted data are still possible. The fault injection mechanism that is introduced in the subsequent commit may help catch more of them. buf_page_import_corrupt_failure: Remove the fault injection, and instead corrupt some pages using Perl code in the tests. btr_cur_pessimistic_insert(): Always reserve extents (except for the change buffer), in order to prevent a subsequent allocation failure. btr_pcur_open_at_rnd_pos(): Merged to the only caller ibuf_merge_pages(). btr_assert_not_corrupted(), btr_corruption_report(): Remove. Similar checks are already part of btr_block_get(). FSEG_MAGIC_N_BYTES: Replaces FSEG_MAGIC_N_VALUE. dict_hdr_get(), trx_rsegf_get_new(), trx_undo_page_get(), trx_undo_page_get_s_latched(): Replaced with error-checking calls. trx_rseg_t::get(mtr_t*): Replaces trx_rsegf_get(). trx_rseg_header_create(): Let the caller update the TRX_SYS page if needed. trx_sys_create_sys_pages(): Merged with trx_sysf_create(). dict_check_tablespaces_and_store_max_id(): Do not access DICT_HDR_MAX_SPACE_ID, because it was already recovered in dict_boot(). Merge dict_check_sys_tables() with this function. dir_pathname(): Replaces os_file_make_new_pathname(). row_undo_ins_remove_sec(): Do not modify the undo page by adding a terminating NUL byte to the record. btr_decryption_failed(): Report decryption failures dict_set_corrupted_by_space(), dict_set_encrypted_by_space(), dict_set_corrupted_index_cache_only(): Remove. dict_set_corrupted(): Remove the constant parameter dict_locked=false. Never flag the clustered index corrupted in SYS_INDEXES, because that would deny further access to the table. It might be possible to repair the table by executing ALTER TABLE or OPTIMIZE TABLE, in case no B-tree leaf page is corrupted. dict_table_skip_corrupt_index(), dict_table_next_uncorrupted_index(), row_purge_skip_uncommitted_virtual_index(): Remove, and refactor the callers to read dict_index_t::type only once. dict_table_is_corrupted(): Remove. dict_index_t::is_btree(): Determine if the index is a valid B-tree. BUF_GET_NO_LATCH, BUF_EVICT_IF_IN_POOL: Remove. UNIV_BTR_DEBUG: Remove. Any inconsistency will no longer trigger assertion failures, but error codes being returned. buf_corrupt_page_release(): Replaced with a direct call to buf_pool.corrupted_evict(). fil_invalid_page_access_msg(): Never crash on an invalid read; let the caller of buf_page_get_gen() decide. btr_pcur_t::restore_position(): Propagate failure status to the caller by returning CORRUPTED. opt_search_plan_for_table(): Simplify the code. row_purge_del_mark(), row_purge_upd_exist_or_extern_func(), row_undo_ins_remove_sec_rec(), row_undo_mod_upd_del_sec(), row_undo_mod_del_mark_sec(): Avoid mem_heap_create()/mem_heap_free() when no secondary indexes exist. row_undo_mod_upd_exist_sec(): Simplify the code. row_upd_clust_step(), dict_load_table_one(): Return DB_TABLE_CORRUPT if the clustered index (and therefore the table) is corrupted, similar to what we do in row_insert_for_mysql(). fut_get_ptr(): Replace with buf_page_get_gen() calls. buf_page_get_gen(): Return nullptr and *err=DB_CORRUPTION if the page is marked as freed. For other modes than BUF_GET_POSSIBLY_FREED or BUF_PEEK_IF_IN_POOL this will trigger a debug assertion failure. For BUF_GET_POSSIBLY_FREED, we will return nullptr for freed pages, so that the callers can be simplified. The purge of transaction history will be a new user of BUF_GET_POSSIBLY_FREED, to avoid crashes on corrupted data. buf_page_get_low(): Never crash on a corrupted page, but simply return nullptr. fseg_page_is_allocated(): Replaces fseg_page_is_free(). fts_drop_common_tables(): Return an error if the transaction was rolled back. fil_space_t::set_corrupted(): Report a tablespace as corrupted if it was not reported already. fil_space_t::io(): Invoke fil_space_t::set_corrupted() to report out-of-bounds page access or other errors. Clean up mtr_t::page_lock() buf_page_get_low(): Validate the page identifier (to check for recently read corrupted pages) after acquiring the page latch. buf_page_t::read_complete(): Flag uninitialized (all-zero) pages with DB_FAIL. Return DB_PAGE_CORRUPTED on page number mismatch. mtr_t::defer_drop_ahi(): Renamed from mtr_defer_drop_ahi(). recv_sys_t::free_corrupted_page(): Only set_corrupt_fs() if any log records exist for the page. We do not mind if read-ahead produces corrupted (or all-zero) pages that were not actually needed during recovery. recv_recover_page(): Return whether the operation succeeded. recv_sys_t::recover_low(): Simplify the logic. Check for recovery error. Thanks to Matthias Leich for testing this extensively and to the authors of https://rr-project.org for making it easy to diagnose and fix any failures that were found during the testing.
2022-06-06 13:03:22 +02:00
@param ctx context (for error log reporting) */
void dict_set_corrupted(dict_index_t *index, const char *ctx)
MDEV-25919 preparation: Various cleanup que_eval_sql(): Remove the parameter lock_dict. The only caller with lock_dict=true was dict_stats_exec_sql(), which will now explicitly invoke dict_sys.lock() and dict_sys.unlock() by itself. row_import_cleanup(): Do not unnecessarily lock the dictionary. Concurrent access to the table during ALTER TABLE...IMPORT TABLESPACE is prevented by MDL and the fact that there cannot exist any undo log or change buffer records that would refer to the table or tablespace. row_import_for_mysql(): Do not unnecessarily lock the dictionary while accessing fil_system. Thanks to MDL_EXCLUSIVE that was acquired by the SQL layer, only one IMPORT may be in effect for the table name. row_quiesce_set_state(): Do not unnecessarily lock the dictionary. The dict_table_t::quiesce state is documented to be protected by all index latches, which we are acquiring. dict_table_close(): Introduce a simpler variant with fewer parameters. dict_table_close(): Reduce the amount of calls. We can simply invoke dict_table_t::release() on startup or in DDL operations, or when the table is inaccessible. In none of these cases, there is no need to invalidate the InnoDB persistent statistics. pars_info_t::graph_owns_us: Remove (unused). pars_info_free(): Define inline. fts_delete(), trx_t::evict_table(), row_prebuilt_free(), row_rename_table_for_mysql(): Simplify. row_mysql_lock_data_dictionary(): Remove some references; use dict_sys.lock() and dict_sys.unlock() instead. row_mysql_lock_table(): Remove. Use lock_table_for_trx() instead. ha_innobase::check_if_supported_inplace_alter(), row_create_table_for_mysql(): Simply assert dict_sys.sys_tables_exist(). In commit 49e2c8f0a6fefdeac50925f758090d6bd099768d and commit 1bd681c8b3c5213ce1f7976940a7dc38b48a0d39 srv_start() actually guarantees that the system tables will exist, or the server is in read-only mode, or startup will fail. Reviewed by: Thirunarayanan Balathandayuthapani
2021-08-31 12:54:20 +02:00
ATTRIBUTE_COLD __attribute__((nonnull));
/** Sets merge_threshold in the SYS_INDEXES
@param[in,out] index index
@param[in] merge_threshold value to set */
void
dict_index_set_merge_threshold(
dict_index_t* index,
ulint merge_threshold);
#ifdef UNIV_DEBUG
/** Sets merge_threshold for all indexes in dictionary cache for debug.
@param[in] merge_threshold_all value to set for all indexes */
void
dict_set_merge_threshold_all_debug(
uint merge_threshold_all);
#endif /* UNIV_DEBUG */
/** Validate the table flags.
@param[in] flags Table flags
@return true if valid. */
UNIV_INLINE
bool
dict_tf_is_valid(
ulint flags);
/** Validate both table flags and table flags2 and make sure they
are compatible.
@param[in] flags Table flags
@param[in] flags2 Table flags2
@return true if valid. */
UNIV_INLINE
bool
dict_tf2_is_valid(
ulint flags,
ulint flags2);
/*********************************************************************//**
This function should be called whenever a page is successfully
compressed. Updates the compression padding information. */
void
dict_index_zip_success(
/*===================*/
dict_index_t* index) /*!< in/out: index to be updated. */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/*********************************************************************//**
This function should be called whenever a page compression attempt
fails. Updates the compression padding information. */
void
dict_index_zip_failure(
/*===================*/
dict_index_t* index) /*!< in/out: index to be updated. */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull));
/*********************************************************************//**
Return the optimal page size, for which page will likely compress.
@return page size beyond which page may not compress*/
ulint
dict_index_zip_pad_optimal_page_size(
/*=================================*/
dict_index_t* index) /*!< in: index for which page size
is requested */
2016-06-21 14:21:03 +02:00
MY_ATTRIBUTE((nonnull, warn_unused_result));
/*************************************************************//**
Convert table flag to row format string.
@return row format name */
const char*
dict_tf_to_row_format_string(
/*=========================*/
ulint table_flag); /*!< in: row format setting */
/** encode number of columns and number of virtual columns in one
4 bytes value. We could do this because the number of columns in
InnoDB is limited to 1017
@param[in] n_col number of non-virtual column
@param[in] n_v_col number of virtual column
@return encoded value */
UNIV_INLINE
ulint
dict_table_encode_n_col(
ulint n_col,
ulint n_v_col);
/** Decode number of virtual and non-virtual columns in one 4 bytes value.
@param[in] encoded encoded value
@param[in,out] n_col number of non-virtual column
@param[in,out] n_v_col number of virtual column */
UNIV_INLINE
void
dict_table_decode_n_col(
ulint encoded,
ulint* n_col,
ulint* n_v_col);
/** Free the virtual column template
@param[in,out] vc_templ virtual column template */
UNIV_INLINE
void
dict_free_vc_templ(
dict_vcol_templ_t* vc_templ);
/** Check whether the table have virtual index.
@param[in] table InnoDB table
@return true if the table have virtual index, false otherwise. */
UNIV_INLINE
bool
dict_table_have_virtual_index(
dict_table_t* table);
2022-01-13 17:27:28 +01:00
#include "dict0dict.inl"
#endif