Home - Waterfall Grid T-Grid Console Builders Recent Builds Buildslaves Changesources - JSON API - About

Console View


Categories: connectors experimental galera main
Legend:   Passed Failed Warnings Failed Again Running Exception Offline No data

connectors experimental galera main
f8c3ac59e964...
f8c3ac59e96429c6af47dffb9a5dd044599a99c6
Sergei Golubchik
MDEV-41496 OOB write with RPAD() crash and multi-byte charsets and uservar

item's charset is metadata, it's known at fix_fields and does not
change later. user variables can violate this invariant if they're
changed and used in the same query.

Let's always wrap such variables in Item_func_conv_charset()
to make sure the caller always gets the same fixed charset, no matter
what the actual variable value is.

Variables that aren't changed in the same query do not get this
unnecessary conversion.
ea82cee30b8f...
ea82cee30b8f999afa10ad7409cf17ef324c6279
Marko Mäkelä
Merge 11.4 into 11.8
e810e919dd73...
e810e919dd73eeb8541eb4544c8c51a6ad1a23cd
Marko Mäkelä
MDEV-41565 Better innodb_log_recovery_target check

log_parse_start(): Account for the mini-transaction trailer.

Also, fix the spelling: fulfil, not "fulfill".

Reviewed by: Alessandro Vetere
e55411904543...
e55411904543b65e49c1cc3fca1138dca9cd3a1b
Teemu Ollakka
MDEV-40727 rsync SST donor stalls in a whole-host sync()

The rsync SST donor ran a bare sync() between the table flush and the
transfer. sync() flushes every mounted filesystem on the host, so its
duration depends on I/O unrelated to the SST. The donor holds FTWRL,
LOCK_log and a paused provider at that point, so the whole server is
frozen until sync() returns.

Flush only the filesystems that hold the files to transfer: the data
directory, an absolute binlog directory, and each symlinked database
directory.

Use the full sync() where 'sync -f' (syncfs) is not available.
e1487d080909...
e1487d080909e6e40f46e0c5d8b2f315dab24587
Aleksey Midenkov
MDEV-40821 Wrong result after PORTION OF with subquery (11.4)

Single-table DELETE converted to multi-table (when a subquery
references the target table) skipped FOR PORTION OF row splitting and
the constant expression check, so it deleted whole rows instead of the
portion.

The fix prohibits processing DELETE ... FOR PORTION OF as multi-table,
like DELETE ... RETURNING.
df75f3d906dc...
df75f3d906dcda25112182be6bd1fa77a30af25d
Sergei Golubchik
CONNECT udfs crash on long filename

namely, json_get_item, jsonget_string, json_set_item, jbin_set_item,
bson_get_item, bsonget_string, bson_set_item

Assisted-by: Claude:claude-opus-5-5
dd47e6279cb6...
dd47e6279cb67790d08c3ac575b549ef206462a9
Sergei Golubchik
jfile_bjson/jfile_convert/bfile_bjson/bfile_convert crash on lrecl=NULL

tests for
* constant NULL
* not constant NULL
* not constant not NULL

Assisted-by: Claude:claude-opus-5-5
d78ca74040f8...
d78ca74040f8048fb98aea6ef1b5576d99b1d0d5
Aleksey Midenkov
MDEV-40821 Wrong result after DELETE HISTORY or PORTION OF (10.11)

The parser assigned versioning conditions of DELETE HISTORY ... BEFORE
SYSTEM_TIME and period conditions of DELETE ... FOR PORTION OF to the
table pointed by last_table(). When the BEFORE SYSTEM_TIME or FOR
PORTION OF expression references another table (e.g. via a subquery or
a sequence function), last_table() is already switched to that table,
and the DELETE target gets no conditions.

DELETE HISTORY then deletes current rows permanently instead of
history, keeps the history it had to purge, and is checked for DELETE
privilege instead of DELETE HISTORY. DELETE HISTORY from a
non-versioned table is not rejected with ER_VERS_NOT_VERSIONED.
DELETE ... FOR PORTION OF deletes all rows of the table, ignoring the
period.

The fix assigns the conditions to the TABLE_LIST returned by
delete_single_table, i.e. the DELETE target itself, regardless of
tables added later in the statement.

DBUG_ASSERT in TABLE::vers_switch_partition() is removed. It checked
delete_history on last_table() and assumed no other table is opened
for write by DELETE HISTORY, which is false for the underlying table
of a view or a table locked by SELECT ... FOR UPDATE in the history
point. The DELETE HISTORY target is already skipped there by its own
delete_history check.
d0dba1f4f02b...
d0dba1f4f02ba3de38358a15d36291a952c707cf
Aleksey Midenkov
MDEV-25529 Auto-create: Pre-existing historical data is not partitioned as specified by ALTER

Adds logic into prep_alter_part_table() for INTERVAL ... AUTO to read the
history range (vers_get_history_range()) and size the new layout from it.
STARTS is aligned to the oldest history row_end (vers_set_starts()) if it
was not specified by the user, or if the user specified it later than
that; in the latter case WARN_VERS_WRONG_STARTS is printed. The history
partitions cover both the newest history row_end and the query time, so
DML after the ALTER does not need to create partitions. A PARTITIONS N
clause is kept if it gives more partitions. An explicit subpartition list
is replaced by as many default subpartitions. If the layout needs more
than MAX_PARTITIONS partitions (counting subpartitions), or the INTERVAL
is longer than the date range, the ALTER fails with
ER_VERS_HIST_PART_FAILED. Month and year intervals are stepped in UTC; a
boundary past the maximum TIMESTAMP ends a partial last history
partition.

In case of fast ALTER TABLE, f.ex. when partitioning already exists, the
above logic is ignored unless FORCE clause is specified. When user
specifies partition list explicitly the above logic is ignored even with
FORCE clause. With FORCE on a table already partitioned by SYSTEM_TIME
the work is done on a clone of the partitioning info. The parsed settings
are copied into the clone's vers_info with
Vers_part_info::copy_settings(), which keeps the clone's now_part and
hist_part (the parsed ones are NULL).

Before reading the range, the metadata lock is upgraded to
MDL_SHARED_NO_WRITE, and then the copy does not run online, so DML
cannot add history that would not fit the computed layout until the
ALTER ends (Alter_info::vers_history_read). LOCK=NONE and
ALGORITHM=INPLACE/NOCOPY/INSTANT are rejected for such ALTER before the
upgrade, so they do not wait for concurrent DML; the ALGORITHM error is
reported by the new Alter_info::report_algorithm_unsupported(), which
mysql_alter_table() uses as well. The range is read under a table lock
taken with mysql_lock_tables(), as some engines (Aria) set up their
transaction in THR_LOCK callbacks. The scan can be killed.

WARN_VERS_STARTS_BEYOND_INTERVAL for ALTER .. AUTO is checked after the
alignment, so it is not printed when STARTS was re-aligned to the history.

vers_get_history_range() detects if an index can be used for row_end
min/max stats and if so it gets it with ha_index_first() and
HA_READ_BEFORE_KEY (as it must ignore current data). Otherwise it does a
table scan to read the stats. Current rows with the pre-11.5 row_end
marker are not taken for history either. In the index they sort below
history after 2038, so the minimum is read after them. There is
test_mdev-25529 debug keyword to check the both and compare results.
WARN_VERS_SLOW_ROW_END is printed if the scan read any row and there is no
usable index (a disabled one does not count as missing).

Static key_cmp was renamed to key_eq to resolve compilation after key.h
was included as key_cmp was already declared there.
d0015644dab4...
d0015644dab42d34069323333190c3f7ce08efaa
Sergei Petrunia
Disable cloning for Item_sum_sp, Enable deletion of test clone for Item_func_in
cc1a2250a6a8...
cc1a2250a6a8f39e097fdb63172f7866c25d3d1e
Aleksey Midenkov
MDEV-40821 SIGSEGV in Window_funcs_sort::setup

Window_funcs_sort failed accessing win_func->window_spec as window
spec was not defined in the query. Usually this is checked by fixing
item win_func, but it was not done by vers_setup_conds().

Actually, earlier check by vers_setup_conds() must fail on DELETE
HISTORY from non-versioned table, but TABLE_LIST for t1 has no
versioning conditions.

The cause was the parser assigning versioning conditions to wrong
table pointed by last_table() which was already switched to another
table from SYSTEM_TIME expression (cs1).

The fix assigns versioning conditions to query_tables, as
delete_single_table puts the DELETE target first there, and tables
referenced later in the statement are appended after it.

Same fix applied to delete_single_table_for_period (period_conditions)
and to DBUG_ASSERT in TABLE::vers_switch_partition(), which checked
delete_history on last_table(). Tests added for both cases.

Without a window function the bug does not crash but silently loses
data: when the BEFORE SYSTEM_TIME or FOR PORTION OF expression
references another table (e.g. via a subquery), the conditions go to
that table. DELETE HISTORY then deletes current rows instead of
history, and is checked for DELETE privilege instead of DELETE
HISTORY. DELETE ... FOR PORTION OF deletes all rows of the table,
ignoring the period.

DELETE ... FOR PORTION OF converted to multi-table (due to index hints
or a subquery on the target table) deleted whole rows instead of
splitting them and skipped the constant expression check. The fix
prohibits processing it as multi-table, like DELETE ... RETURNING.

Warning WARN_INDEX_HINTS_IGNORED now names the clause the index hints
are incompatible with: RETURNING or FOR PORTION OF.
c3a21cb661f0...
c3a21cb661f030e11b59c583af1fe7d41336d802
Sergei Golubchik
MDEV-41603 DECODE_HISTOGRAM crashes when decoding JSON_HB

don't return a pointer to a local buffer
bb7cc8db91b6...
bb7cc8db91b65618bf18fba36f35d242e4c3f411
Sergei Golubchik
MDEV-41608 some CONNECT jfile/bfile udfs ignore FILE and secure_file_priv

also they should all have maybe_null=true
b2fa0bd4dbb0...
b2fa0bd4dbb0dab8ae75a4e65d6854c6c45601cf
Sergei Golubchik
fix NULL handling for jfile_bjson/jfile_convert/bfile_bjson/bfile_convert

to be consistent with other functions

Assisted-by: Claude:claude-opus-5-5
ab69e9189db3...
ab69e9189db31d6402efdd8dcd9e0c09c9d89be1
Sergei Golubchik
MDEV-41641 CONNECT REST can bypass FILE protection

only allow http and https schemes for REST
a86e83d6a21e...
a86e83d6a21e5ccd9aa4445e046c4929c19f9b41
Aleksey Midenkov
MDEV-40821 Wrong result after DELETE HISTORY or PORTION OF (10.11)

The parser assigned versioning conditions of DELETE HISTORY ... BEFORE
SYSTEM_TIME and period conditions of DELETE ... FOR PORTION OF to the
table pointed by last_table(). When the BEFORE SYSTEM_TIME or FOR
PORTION OF expression references another table (e.g. via a subquery or
a sequence function), last_table() is already switched to that table,
and the DELETE target gets no conditions.

DELETE HISTORY then deletes current rows permanently instead of
history, keeps the history it had to purge, and is checked for DELETE
privilege instead of DELETE HISTORY. DELETE HISTORY from a
non-versioned table is not rejected with ER_VERS_NOT_VERSIONED.
DELETE ... FOR PORTION OF deletes all rows of the table, ignoring the
period.

The fix assigns the conditions to the TABLE_LIST returned by
delete_single_table, i.e. the DELETE target itself, regardless of
tables added later in the statement.

DBUG_ASSERT in TABLE::vers_switch_partition() is removed. It checked
delete_history on last_table() and assumed no other table is opened
for write by DELETE HISTORY, which is false for the underlying table
of a view or a table locked by SELECT ... FOR UPDATE in the history
point. The DELETE HISTORY target is already skipped there by its own
delete_history check.
a38fc9d8baa5...
a38fc9d8baa599b819606b58565df84b17f089a3
Aleksey Midenkov
MDEV-40821 Wrong result after PORTION OF with subquery (11.4)

Single-table DELETE converted to multi-table (when a subquery
references the target table) skipped FOR PORTION OF row splitting and
the constant expression check, so it deleted whole rows instead of the
portion.

The fix prohibits processing DELETE ... FOR PORTION OF as multi-table,
like DELETE ... RETURNING.
958a290cddf8...
958a290cddf818c40336a161162859e682feb8b7
Aleksey Midenkov
MDEV-40821 Wrong result after PORTION OF with subquery (11.4)

Single-table DELETE converted to multi-table (when a subquery
references the target table) skipped FOR PORTION OF row splitting and
the constant expression check, so it deleted whole rows instead of the
portion.

The fix prohibits processing DELETE ... FOR PORTION OF as multi-table,
like DELETE ... RETURNING.
9543e559d4f3...
9543e559d4f3c4db1a8774c6e6881790ff747933
Sergei Golubchik
fix RPAD(), LPAD() and INSERT() not to mangle its arguments

they used to overwrite the charset of their arguments, this is
incorrect and led to wrong results
93c0135a95f4...
93c0135a95f451089ccc1a86747ca150bebed50d
Sergei Golubchik
CONNECT: more FILE and secure_file_priv checks in udfs

Assisted-by: Claude:claude-opus-5-5
8daa8c0d0f62...
8daa8c0d0f62807878ca6471ab49d137b8c6d376
Sergei Golubchik
cleanup: connect.rest test

no need to run the same CREATE TABLE twice, first to check if it works
and then to test its result
82b736f04eda...
82b736f04eda7cb0d3ccffb1f79f167865184baf
mariadb-PranavTiwari
Added sci
7d5ffe0f4f7c...
7d5ffe0f4f7cfbc62432fd99184099bd5cbd78a9
Sergei Golubchik
MDEV-41613 CONNECT jfile_convert crashes on invalid file

Assisted-by: Claude:claude-opus-5-5
7cbb938f48f6...
7cbb938f48f6fd0af597329661f9672cc37b6d67
Sergei Golubchik
cleanup: remove unused variable
72dbf7192f90...
72dbf7192f907fa377d0937ff8e125e4b62daaf4
Aleksey Midenkov
MDEV-40821 Wrong result after DELETE HISTORY or PORTION OF (10.11)

The parser assigned versioning conditions of DELETE HISTORY ... BEFORE
SYSTEM_TIME and period conditions of DELETE ... FOR PORTION OF to the
table pointed by last_table(). When the BEFORE SYSTEM_TIME or FOR
PORTION OF expression references another table (e.g. via a subquery or
a sequence function), last_table() is already switched to that table,
and the DELETE target gets no conditions.

DELETE HISTORY then deletes current rows permanently instead of
history, keeps the history it had to purge, and is checked for DELETE
privilege instead of DELETE HISTORY. DELETE HISTORY from a
non-versioned table is not rejected with ER_VERS_NOT_VERSIONED.
DELETE ... FOR PORTION OF deletes all rows of the table, ignoring the
period.

The fix assigns the conditions to the TABLE_LIST returned by
delete_single_table, i.e. the DELETE target itself, regardless of
tables added later in the statement.

DBUG_ASSERT in TABLE::vers_switch_partition() is removed. It checked
delete_history on last_table() and assumed no other table is opened
for write by DELETE HISTORY, which is false for the underlying table
of a view or a table locked by SELECT ... FOR UPDATE in the history
point. The DELETE HISTORY target is already skipped there by its own
delete_history check.
70c3641d9c5b...
70c3641d9c5b1a1284bf2359cfebf59cfd51d340
Sergei Golubchik
Item_func_decode_histogram::val_str shouldn't modify its arguments
55a21a72a6ad...
55a21a72a6ad0220018db3e024812a746c85088a
Aleksey Midenkov
MDEV-40821 Wrong result after DELETE HISTORY or PORTION OF (10.11)

The parser assigned versioning conditions of DELETE HISTORY ... BEFORE
SYSTEM_TIME and period conditions of DELETE ... FOR PORTION OF to the
table pointed by last_table(). When the BEFORE SYSTEM_TIME or FOR
PORTION OF expression references another table (e.g. via a subquery or
a sequence function), last_table() is already switched to that table,
and the DELETE target gets no conditions.

DELETE HISTORY then deletes current rows permanently instead of
history, keeps the history it had to purge, and is checked for DELETE
privilege instead of DELETE HISTORY. DELETE HISTORY from a
non-versioned table is not rejected with ER_VERS_NOT_VERSIONED.
DELETE ... FOR PORTION OF deletes all rows of the table, ignoring the
period.

The fix assigns the conditions to the TABLE_LIST returned by
delete_single_table, i.e. the DELETE target itself, regardless of
tables added later in the statement.

DBUG_ASSERT in TABLE::vers_switch_partition() is removed. It checked
delete_history on last_table() and assumed no other table is opened
for write by DELETE HISTORY, which is false for the underlying table
of a view or a table locked by SELECT ... FOR UPDATE in the history
point. The DELETE HISTORY target is already skipped there by its own
delete_history check.
4d3851d69c2c...
4d3851d69c2c77c775a89ff031554d58e58c3798
Aleksey Midenkov
MDEV-40821 Wrong result after DELETE HISTORY or PORTION OF (10.11)

The parser assigned versioning conditions of DELETE HISTORY ... BEFORE
SYSTEM_TIME and period conditions of DELETE ... FOR PORTION OF to the
table pointed by last_table(). When the BEFORE SYSTEM_TIME or FOR
PORTION OF expression references another table (e.g. via a subquery or
a sequence function), last_table() is already switched to that table,
and the DELETE target gets no conditions.

DELETE HISTORY then deletes current rows permanently instead of
history, keeps the history it had to purge, and is checked for DELETE
privilege instead of DELETE HISTORY. DELETE HISTORY from a
non-versioned table is not rejected with ER_VERS_NOT_VERSIONED.
DELETE ... FOR PORTION OF deletes all rows of the table, ignoring the
period.

The fix assigns the conditions to the TABLE_LIST returned by
delete_single_table, i.e. the DELETE target itself, regardless of
tables added later in the statement.

DBUG_ASSERT in TABLE::vers_switch_partition() is removed. It checked
delete_history on last_table() and assumed no other table is opened
for write by DELETE HISTORY, which is false for the underlying table
of a view or a table locked by SELECT ... FOR UPDATE in the history
point. The DELETE HISTORY target is already skipped there by its own
delete_history check.
48d70262c107...
48d70262c1077ed2d6e97813f4ef92c145fd4660
Jan Lindström
MDEV-40727 rsync SST symlink coverage gaps

Failure:
- Filesystem flush before transfer only checks symlinks directly under
  the data directory, missing a per-table DATA/INDEX DIRECTORY symlink
  one level deeper (MyISAM, Aria), and missing InnoDB's per-table DATA
  DIRECTORY entirely, since that uses a '.isl' text file, not a symlink.
- The database directory transfer recreates symlinks as symlinks. The
  joiner's rsync daemon cannot keep an absolute symlink target as-is,
  so a MyISAM or Aria table placed by DATA DIRECTORY or INDEX DIRECTORY
  comes out unreadable after the SST.

Fix:
- Extend the flush's symlink search to two levels deep, and add a scan
  for '.isl' files, flushing the filesystem each one points to.
- Transfer the database directories with --copy-links instead of
  --links, so a table file reached through a symlink is copied by
  content instead of recreated as a broken symlink.

Test:
- MDEV-40727.test creates an Aria table with DATA DIRECTORY and INDEX
  DIRECTORY and an InnoDB table with DATA DIRECTORY, forces a full SST,
  then checks the donor's log for the flush of both external
  directories and checksums every table in database test between the
  two nodes.

Assisted-by: https://mariadb.org/governance/governance-ai-policy/
3fea08c51c3a...
3fea08c51c3ae79f8bad754fcc1ba3d7bf717f53
Aleksey Midenkov
MDEV-40821 SIGSEGV in Window_funcs_sort::setup

Window_funcs_sort failed accessing win_func->window_spec as window
spec was not defined in the query. Usually this is checked by fixing
item win_func, but it was not done by vers_setup_conds().

Actually, earlier check by vers_setup_conds() must fail on DELETE
HISTORY from non-versioned table, but TABLE_LIST for t1 has no
versioning conditions.

The cause was the parser assigning versioning conditions to wrong
table pointed by last_table() which was already switched to another
table from SYSTEM_TIME expression (cs1).

The fix assigns versioning conditions to query_tables, as
delete_single_table puts the DELETE target first there, and tables
referenced later in the statement are appended after it.

Same fix applied to delete_single_table_for_period (period_conditions)
and to DBUG_ASSERT in TABLE::vers_switch_partition(), which checked
delete_history on last_table(). Tests added for both cases.

Without a window function the bug does not crash but silently loses
data: when the BEFORE SYSTEM_TIME or FOR PORTION OF expression
references another table (e.g. via a subquery), the conditions go to
that table. DELETE HISTORY then deletes current rows instead of
history, and is checked for DELETE privilege instead of DELETE
HISTORY. DELETE ... FOR PORTION OF deletes all rows of the table,
ignoring the period.

DELETE ... FOR PORTION OF converted to multi-table (due to index hints
or a subquery on the target table) deleted whole rows instead of
splitting them and skipped the constant expression check. The fix
prohibits processing it as multi-table, like DELETE ... RETURNING.

Warning WARN_INDEX_HINTS_IGNORED now names the clause the index hints
are incompatible with: RETURNING or FOR PORTION OF.
3e76b3762cf9...
3e76b3762cf9adccfcec09eab8bd251b0d3ebc6d
Aleksey Midenkov
MDEV-40821 Wrong result after PORTION OF with subquery (11.4)

Single-table DELETE converted to multi-table (when a subquery
references the target table) skipped FOR PORTION OF row splitting and
the constant expression check, so it deleted whole rows instead of the
portion.

The fix prohibits processing DELETE ... FOR PORTION OF as multi-table,
like DELETE ... RETURNING.
38f50ad36642...
38f50ad36642a16d8a61a3f7adbb1a8cfe747028
Sergei Golubchik
MDEV-41626 definer check in CREATE PROCEDURE is too lax
32d918f7daad...
32d918f7daada47fa2174861996f39b6498ca630
Sergei Golubchik
cleanup: use a local HTTP server in connect.rest

The test depended on an external site, so it was slow and unreliable.
Serve std_data/bios.json from a forked perl server instead.

Assisted-by: Claude:claude-opus-5-5
31afbc4ad522...
31afbc4ad5223432307865be55d99109f602805a
mariadb-PranavTiwari
Added sci
2a96bc43852a...
2a96bc43852ad8988cee65118128588fa42c7055
Aleksey Midenkov
MDEV-40799 Runtime plugin/UDF load errors lost under --silent-startup

Regression from MDEV-32745 (7828fb475b0), which guarded the
plugin-load my_error() calls with opt_silent_startup.  That option is
a lifetime global, set once at startup and never reset, so the guard
suppressed the SQL error for the whole server lifetime, not just
during startup.  Runtime operations (INSTALL PLUGIN, CREATE FUNCTION
... SONAME) then skipped my_error(), never set the diagnostics area
and wrongly succeeded - e.g. main.ps's "call proc_1()" no longer
failed with ER_CANT_OPEN_LIBRARY.

Startup callers pass MYF(ME_ERROR_LOG); runtime callers pass MYF(0).
Gate the silencing on that flag via silent_plugin_startup() so it
applies only to the startup error-log path, and runtime errors always
reach the client.

No new test case: the runtime failure path is already covered by
existing tests (e.g. main.ps's ER_CANT_OPEN_LIBRARY check).  The
regression stayed invisible only because stock MTR does not start
servers with --silent-startup.  A dedicated test would have to restart
the server with --silent-startup solely to assert that a startup-only
option does not affect runtime, which adds little over the restored
invariant.
268c295ec1e1...
268c295ec1e15b920cc66902dace072230624c87
Sergei Golubchik
MDEV-41622 INSERT crashes on deeply nested GeometryCollection

- get_data_size() and init_from_wkb() return their error values on
  stack overrun, not a valid-looking length
- add stack checks to the remaining recursive GeometryCollection
  methods: wkt, json, mbr, area, length, dimension, store_shapes
- test non-GEOMETRY column and SELECT paths

Assisted-by: Claude:claude-opus-5-5
221134b73497...
221134b73497a01b8fc8c5af6e203a77db4882bf
Sergei Golubchik
cleanup: use StringBuffer<>
10b5d7f15744...
10b5d7f15744ce728d04fa748a9c6b8f047e48f7
Aleksey Midenkov
MDEV-25529 Auto-create: Pre-existing historical data is not partitioned as specified by ALTER

Adds logic into prep_alter_part_table() for INTERVAL ... AUTO to read the
history range (vers_get_history_range()) and size the new layout from it.
STARTS is aligned to the oldest history row_end (vers_set_starts()) if it
was not specified by the user, or if the user specified it later than
that; in the latter case WARN_VERS_WRONG_STARTS is printed. The history
partitions cover both the newest history row_end and the query time, so
DML after the ALTER does not need to create partitions. A PARTITIONS N
clause is kept if it gives more partitions. An explicit subpartition list
is replaced by as many default subpartitions. If the layout needs more
than MAX_PARTITIONS partitions (counting subpartitions), or the INTERVAL
is longer than the date range, the ALTER fails with
ER_VERS_HIST_PART_FAILED. Month and year intervals are stepped in UTC; a
boundary past the maximum TIMESTAMP ends a partial last history
partition.

In case of fast ALTER TABLE, f.ex. when partitioning already exists, the
above logic is ignored unless FORCE clause is specified. When user
specifies partition list explicitly the above logic is ignored even with
FORCE clause. With FORCE on a table already partitioned by SYSTEM_TIME
the work is done on a clone of the partitioning info. The parsed settings
are copied into the clone's vers_info with
Vers_part_info::copy_settings(), which keeps the clone's now_part and
hist_part (the parsed ones are NULL).

Before reading the range, the metadata lock is upgraded to
MDL_SHARED_NO_WRITE, and then the copy does not run online, so DML
cannot add history that would not fit the computed layout until the
ALTER ends (Alter_info::vers_history_read). LOCK=NONE and
ALGORITHM=INPLACE/NOCOPY/INSTANT are rejected for such ALTER before the
upgrade, so they do not wait for concurrent DML; the ALGORITHM error is
reported by the new Alter_info::report_algorithm_unsupported(), which
mysql_alter_table() uses as well. The range is read under a table lock
taken with mysql_lock_tables(), as some engines (Aria) set up their
transaction in THR_LOCK callbacks. The scan can be killed.

WARN_VERS_STARTS_BEYOND_INTERVAL for ALTER .. AUTO is checked after the
alignment, so it is not printed when STARTS was re-aligned to the history.

vers_get_history_range() detects if an index can be used for row_end
min/max stats and if so it gets it with ha_index_first() and
HA_READ_BEFORE_KEY (as it must ignore current data). Otherwise it does a
table scan to read the stats. Current rows with the pre-11.5 row_end
marker are not taken for history either. In the index they sort below
history after 2038, so the minimum is read after them. There is
test_mdev-25529 debug keyword to check the both and compare results.
WARN_VERS_SLOW_ROW_END is printed if the scan read any row and there is no
usable index (a disabled one does not count as missing).

Static key_cmp was renamed to key_eq to resolve compilation after key.h
was included as key_cmp was already declared there.
05377d6c9bf3...
05377d6c9bf3458245c042e828e4373f8a18d690
Sergei Golubchik
cleanup: connect test
02a88214b7eb...
02a88214b7eb9895239e92a2711dd3cb0ddc06c5
forkfun
MDEV-24557 mariadb-dump: import MySQL 5.7 and 8.x users, grants, roles

mariadb-dump --system=users/all now dumps a MySQL 5.7+ source as
MariaDB statements. MySQL is recognised by the version string and the
absence of mysql.roles_mapping, so --version= cannot fool it.

- Privileges are mapped to MariaDB ones; others are dropped with a
  "# WARNING". WITH GRANT OPTION is dropped from a mapped privilege.
- CREATE USER clauses MariaDB lacks are dropped, with a warning for
  the lossy ones. An unsupported auth plugin locks the account.
- Roles are inferred from mysql.user, since MySQL has no role flag.
  Same-named roles on several hosts merge into one.
- A statement containing "*/" is emitted as EXECUTE IMMEDIATE
  CONCAT('..*', '/..'), so it cannot end the /*M!...*/ wrapper.
- The self-DROP guard now matches current_user() and runs before the
  first DROP USER.

Translators are in client/mysql8_translate.cc. The MySQL-source paths
run only against a real MySQL; recapture mysql8_users_dump*.sql
(recipe in mysql8_users_setup.sql) when they change.

Assisted-By: Claude:claude-5.5-sonnet
BuildBot (0.8.9) working for the MariaDB project.
Page built: Tue 20 Oct 2026 13:10:29 (UTC)