Log message:
mariadb105: Update to 10.5.24

                           MariaDB 10.5.24 Changelog

 * Includes all fixes from MariaDB 10.4.33
  * MDEV-32837 long unique does not work like unique key when using
    replace
  * MDEV-29954 Unique hash key on column prefix is computed
    incorrectly
  * cleanup: MY_STRNNCOLLSP_NCHARS_EMULATE_TRIMMED_TRAILING_SPACES
  * cleanup: unused and undefined methods
  * MDEV-11628 mysql.slow_log reports incorrect start time
  * cleanup: main.log_tables test
  * disable perfschema in mtr bootstrap
  * MDEV-28651 quote(NULL) returns incorrect result in view ('NU'
    instead of 'NULL')
  * MDEV-33290: Disable ColumnStore based on boost version
  * sql_test.cc compile fix
  * MDEV-33165 Incorrect result interceptor passed to
    mysql_explain_union()
  * MDEV-33283: Binlog Checksum is Zeroed by Zlib if Part of Event
    Data is Empty
  * MDEV-32968 InnoDB fails to restore tablespace first page from
    doublewrite buffer when page is empty
  * Replace incorrect message `mariadb-safe` with correct
    `mariadbd-safe`
  * MDEV-27666 User variable not parsed as geometry variable in
    geometry function
  * MDEV-32968 InnoDB fails to restore tablespace first page from
    doublewrite buffer when page is empty
  * MDEV-32111: Debian Sid/Trixie will not have libncurses 5 anymore
  * MDEV-29298 INSERT ... SELECT Does not produce an optimizer trace
  * Fix test failures on s390x in test following
    main.column_compression_rpl
  * MDEV-32235: mysql_json cannot be used on newly created table
  * Revert "MDEV-32235: mysql_json cannot be used on newly created
    table"
  * MDEV-32235: mysql_json cannot be used on newly created table
  * MDEV-32790: Output result in show create table for mysql_json
    type should be longtext
  * MDEV-30170 ha_spider::delete_table() should report table not
    exist
  * MDEV-29002 Spider: remove SPIDER_CONN::loop_check_meraged_last
  * MDEV-33092 Undefined reference to concurrency on Solaris
  * MDEV-32984 Update federated table and column privileges
  * MDEV-33098: Fix some instrumentation for innodb.doublewrite_debug
  * MDEV-33156 Crash on innodb_buf_flush_list_now=ON and
    innodb_force_recovery=6
  * MDEV-33098 The test innodb.doublewrite_debug occasionally fails
    to start up InnoDB
  * MDEV-32549 Cluster inconsistent after SAVEPOINT is rolled back
  * MDEV-31905 GTID inconsistency
  * MDEV-33046 fixup. Do not try to schedule timer without
    dict_stats_start()
  * MDEV-33046 - delete unnecessary synchronization with
    dict_stats_mutex
  * MDEV-24481 : galera_3nodes.galera_vote_rejoin_mysqldump MTR
    failed: mysql_shutdown failed
  * Fix main.type_timestamp: Change 10.10 in "End of 10.10 tests" to
    10.5
  * MDEV-33062 innodb_undo_log_truncate=ON prevents fast shutdown
  * MDEV-33052 MSAN use-of-uninitialized-value in
    buf_read_ahead_linear()
  * MDEV-33009 Server hangs for a long time with
    innodb_undo_log_truncate=ON
  * MDEV-33006 Missing required privilege CONNECTION ADMIN
  * MDEV-32839 LONG UNIQUE gives error when used with REPLACE
  * cleanup: remove innodb-specific code around
    update_auto_increment()
  * cleanup: remove partition-specific code around
    update_auto_increment()
  * [fixup] galera: Fix an accidental logical inversion in a recent
    merge
  * MDEV-29816 rpl.rpl_parallel_29322 occasionally fails in BB
  * MDEV-16264 fixup: Remove a useless test
  * MDEV-17226 Column Data in Truncated on UNION to the length of the
    first value if using REPLACE
  * Cleanup: Removing the unused method
    Type_handler::get_handler_by_cmp_type
  * MDEV-32971 Assertion !recv_sys.is_corrupt_fs() failed on recovery
  * MDEV-31939 Adaptive flush recommendation ignores dirty ratio and
    checkpoint age
  * MDEV-32242 innodb.doublewrite test case always gets skipped
  * [fixup] post-merge spider fixup
  * Followup for fix for MDEV-20169: enable main.partition_innodb
  * MDEV-32725 innodb.import_update_stats accesses uninitialized
    ib_table->stat_n_rows
  * fix for the test (real fixes will be in 10.4)
  * MDEV-32212 DELETE with ORDER BY and semijoin optimization causing
    crash
  * MDEV-32269 InnoDB after ALTER TABLE…IMPORT TABLESPACE may not be
    crash safe
  * MDEV-28682 gcol.gcol_purge contaminates further execution of
    innodb.gap_locks
  * Remove deprication from mariadbd --debug
  * MDEV-32879 Server crash in my_decimal::operator= or unexpected
    ER_DUP_ENTRY upon comparison with INET6 and similar types
  * MDEV-31611: mariadb-setpermission - Can't use string as an ARRAY
    ref while strict refs in use
  * MDEV-32861 InnoDB hangs when running out of I/O slots
  * MDEV-32820 Race condition between trx_purge_free_segment() and
    trx_undo_create()
  * MDEV-22243 type_test.type_test_double fails with 'NUMERIC_SCALE
    NULL'
  * Revert "MDEV-22243 type_test.type_test_double fails with
    'NUMERIC_SCALE NULL'"
  * MDEV-24784 JSON_ARRAYAGG charset issue
  * MDEV-26055: Correct the formula for adaptive flushing
  * MDEV-26055: Improve adaptive flushing
  * MDEV-32757 innodb_undo_log_truncate=ON is not crash safe
  * MDEV-32689: Remove Ubuntu Bionic from 10.5
  * MDEV-32798 innodb_fast_shutdown=0 hang after incomplete startup
  * bump the VERSION
  * galera: cleanup of the lists of disabled tests
  * MDEV-31413: post-fix for 10.5+ (galera_restart_replica test
    failures)
  * MDEV-32737 innodb.log_file_name fails on Assertion `after_apply
    || !(blocks).end in recv_sys_t::clear
  * Fix random test failures in testcase perfschema.mdl_func
  * MDEV-32672: Don't hold LOCK_thd_data over commit_ordered
  * MDEV-32462: mysql_upgrade -s still checks for non system tables
  * MDEV-32462: mysql_upgrade -s still checks for non system tables
  * MDEV-26875: Wrong user in SET DEFAULT ROLE error
  * galera: post-fix after migrating changes from 10.4

Log message:
mariadb105: Update to 10.5.23

Fixes various crashes and data corruption issues.

Release notes:
https://mariadb.com/kb/en/mariadb-10-5-23-release-notes/

Log message:
*: recursive bump for icu 74.1

Log message:
*: bump for openssl 3

Log message:
mariadb*-client: Force HAVE_CURSES_H when using ncurses.

For some reason mariadb removed any cmake detection for HAVE_CURSES_H so it is
never defined, but the source is still littered with it, causing compilation
problems because curses.h never gets included.

The whole thing is incredibly messy, and this is the least likely to
accidentally break some alternate configuration, but implementations that use
curses.h which isn't ncurses are likely to still be broken.

Log message:
mariadb: Correct include order for broken curses detection workarounds.

Log message:
mariadb*-client: Ensure the wide ncurses library is picked if the user chose it

Log message:
mariadb105: update to 10.5.21

                         MariaDB 10.5.21 Release Notes

Notable Items

  InnoDB

     * Server crashes in st_join_table::choose_best_splitting (MDEV-31403)
     * Crash with condition pushable into derived and containing outer
       reference (MDEV-31240)
     * InnoDB does not free UNDO after the fix of MDEV-30671 (MDEV-31234)
     * Revert "MDEV-30473 : Do not allow GET_LOCK() / RELEASE_LOCK() in
       cluster"

  Optimizer

     * Crash with condition pushable into derived and containing outer
       reference (MDEV-31403 MDEV-31240)
     * Crash with EXPLAIN EXTENDED for multi-table update of system table
       (MDEV-31224)

                         MariaDB 10.5.20 Release Notes

Notable Items

  InnoDB

     * Crash on ROLLBACK in a ROW_FORMAT=COMPRESSED table (MDEV-30882)
     * UNIQUE USING HASH accepts duplicate entries for tricky collations
       (MDEV-30034)
     * rec_get_offsets() is not optimal (MDEV-30567)
     * Performance regression in fil_space_t::try_to_close() introduced in
       MDEV-23855 (MDEV-30775)
     * InnoDB recovery hangs when buffer pool ran out of memory (MDEV-30551)
     * InnoDB undo log truncation fails to wait for purge of history
       (MDEV-30671
     * Fix miscount of doublewrites by Innodb_data_written (MDEV-31124)

Backup

     * mariadb-backup doesn't utilise innodb-undo-log-directory (if specified
       as a relative path) during copy-back operation (MDEV-28187)
     * mariabackup issues error messages during InnoDB tablespaces export on
       partial backup preparing (MDEV-29050)
     * mariadb-backup does not copy Aria logs if aria_log_dir_path is used
       (MDEV-30968)
     * Race condition between buffer pool flush and log file deletion in
       mariadb-backup --prepare (MDEV-30860)

  Replication

     * Fixed a deadlock on parallel slave involving full image Write event on
       the sequence engine (MDEV-29621)
     * Fixed an attempted out-of-order binlogging error on slave involving
       ALTER on the sequence engine (MDEV-31077)
     * Corrected non-versioned master to versioned slave replication on
       no-unique attribute table (MDEV-30430)
     * Mended encrypted binlog master to error out to gtid-mode slave when
       master could not decrypt a binlog file (MDEV-28798)
     * Refined optimistic parallel slave to error-exit without any hang
       (MDEV-30780)
     * Ensured SHOW-SLAVE-STATUS is processed on the parallel slave having a
       necessary mutex always intialized (MDEV-30620)
     * Fixed the slave applier to report a correct error when gtid_slave_pos
       insert fails for some (engine) reasons (MDEV-31038)

  Optimizer

     * Split Materialized optimization is improved to re-fill the
       materialized table only if necessary. The fewer number of table
       refills is taken into account when choosing query plan, too
       (MDEV-26301).
     * Queries using SELECT DISTINCT some_expression(aggregate_function())
       could produce wrong query result. (MDEV-20057)
     * A GROUP BY query with MIN(primary_key) in select list and
       primary_key<>const in the WHERE could produce wrong result when
       executed with "Using index for group-by" strategy (MDEV-30605)
     * EXPLAIN could erroneously report that Rowid Filter optimization is
       used for partitioned tables. Partitioned tables do not support it.
       (MDEV-30596)
     * A bug in selectivity computations for SINGLE/DOUBLE_PREC_HB histograms
       could cause wrong estimates to be produced. This could cause the
       optimizer to pick sub-optimal query plans (MDEV-31067).

  Security

     * Fixes for the following security vulnerabilities:
          * CVE-2022-47015