- 23 Jan, 2024 7 commits
-
-
Sergei Golubchik authored
-
Sergei Golubchik authored
use thd->start_time for the "start_time" column of the slow_log table. "current_time" here refers to the current_time() function return value not to the actual *current* time. also fixes MDEV-33267 User with minimal permissions can intentionally corrupt mysql.slow_log table
-
Sergei Golubchik authored
-
Sergei Golubchik authored
should fix numerous failures of main.bootstrap tets
-
Alexander Barkov authored
Item_func_quote did not calculate its max_length correctly for nullable arguments. Fix: In case if the argument is nullable, reserve at least 4 characters so the string "NULL" fits.
-
Daniel Black authored
MCOL-5611 supporting with Boost-1.80, the version "next_prime" disappears from https://github.com/boostorg/unordered/blob/boost-1.79.0/include/boost/unordered/detail/implementation.hpp makes it the currenly highest supported versions. Lets check this version. While CMake-3.19+ supports version ranges in package determinations this isn't supported for Boost in Cmake-3.28. So we check for the 1.80 and don't compile ColumnStore.
-
Dmitry Shulga authored
-
- 22 Jan, 2024 2 commits
-
-
Rex authored
Statements affect by this bug are all SQL statements that 1) prefixed with "EXPLAIN" 2) have a lower level join structure created for a union subquery. A bug in select_describe() passed an incorrect "result" object to mysql_explain_union(), resulting in unpredictable behaviour and out of context calls. Reviewed by: Oleksandr Byelkin, sanja@mariadb.com
-
Brandon Nesterenko authored
An existing binlog checksum can be overridden to 0 if writing a NULL payload when using Zlib for the computation. That is, calling into Zlib's crc32 with empty data initializes an incremental CRC computation to 0. This patch changes the Log_event_writer::write_data() to exit immediately if there is nothing to write, thereby bypassing the checksum computation. This follows the pattern of Log_event_writer::encrypt_and_write(), which also exits immediately if there is no data to write. Reviewed By: ============ Andrei Elkin <andrei.elkin@mariadb.com>
-
- 19 Jan, 2024 1 commit
-
-
Thirunarayanan Balathandayuthapani authored
MDEV-32968 InnoDB fails to restore tablespace first page from doublewrite buffer when page is empty recv_dblwr_t::find_first_page(): Free the allocated memory to read the first 3 pages from tablespace. innodb.doublewrite: Added sleep to ensure page cleaner thread wake up from my_cond_wait
-
- 17 Jan, 2024 1 commit
-
-
Sophist authored
-
- 16 Jan, 2024 1 commit
-
-
Alexander Barkov authored
Adding GEOMETRY type user variables.
-
- 15 Jan, 2024 1 commit
-
-
Thirunarayanan Balathandayuthapani authored
MDEV-32968 InnoDB fails to restore tablespace first page from doublewrite buffer when page is empty - InnoDB fails to find the space id from the page0 of the tablespace. In that case, InnoDB can use doublewrite buffer to recover the page0 and write into the file. - buf_dblwr_t::init_or_load_pages(): Loads only the pages which are valid.(page lsn >= checkpoint). To do that, InnoDB has to open the redo log before system tablespace, read the latest checkpoint information. recv_dblwr_t::find_first_page(): 1) Iterate the doublewrite buffer pages and find the 0th page 2) Read the tablespace flags, space id from the 0th page. 3) Read the 1st, 2nd and 3rd page from tablespace file and compare the space id with the space id which is stored in doublewrite buffer. 4) If it matches then we can write into the file. 5) Return space which matches the pages from the file. SysTablespace::read_lsn_and_check_flags(): Remove the retry logic for validating the first page. After restoring the first page from doublewrite buffer, assign tablespace flags by reading the first page. recv_recovery_read_max_checkpoint(): Reads the maximum checkpoint information from log file recv_recovery_from_checkpoint_start(): Avoid reading the checkpoint header information from log file Datafile::validate_first_page(): Throw error in case of first page validation fails.
-
- 14 Jan, 2024 1 commit
-
-
Tuukka Pasanen authored
Upstream Debian Sid which will become Debian Trixie (13) have dropped NCurses version 5 and changed dev package name just libncurses-dev
-
- 13 Jan, 2024 1 commit
-
-
Oleg Smirnov authored
Add INSERT ... SELECT to the list of commands that can be traced Approved by Sergei Petrunia (sergey@mariadb.com)
-
- 12 Jan, 2024 3 commits
-
-
Kristian Nielsen authored
The problem is the test is skipped after sourcing include/master-slave.inc. This leaves the slave threads running after the test is skipped, causing a following test to fail during rpl setup. Also rename have_normal_bzip.inc to the more appropriate _zlib. Signed-off-by: Kristian Nielsen <knielsen@knielsen-hq.org>
-
Anel Husakovic authored
- Closes PR #2839 - Usage of `Column_definition_fix_attributes()` suggested by Alexandar Barkov - thanks bar, that is better than hook in server code (reverted 22f3ebe4) - This method is called after parsing the data type: * in `CREATE/ALTER TABLE` * in SP: return data type, parameter data type, variable data type - We want to disallow all these use cases of MYSQL_JSON. - Reviewer: bar@mariadb.com cvicentiu@mariadb.org
-
Anel Husakovic authored
This reverts commit 22f3ebe4.
-
- 11 Jan, 2024 4 commits
-
-
Anel Husakovic authored
Closes PR #2839 Reviewer: cvicentiu@mariadb.org
-
Anel Husakovic authored
- We don't test `json` MySQL tables from `std_data` since the error `ER_TABLE_NEEDS_REBUILD ` is invoked. However MDEV-32235 will override this test after merge, but leave it to show behavior and historical changes. - Closes PR #2833 Reviewer: <cvicentiu@mariadb.org> <serg@mariadb.com>
-
Yuchen Pei authored
All Spider tables are recorded in the system table mysql.spider_tables. Deleting a spider table removes the corresponding rows from the system table, among other things. This patch makes it so that if spider could not find any record in the system table to delete for a given table, it should correctly report that no such Spider table exists.
-
Yuchen Pei authored
The field is assigned but unused, and it causes heap-use-after-free.
-
- 10 Jan, 2024 6 commits
-
-
Sergei Golubchik authored
remove thr_setconcurrency() followup for 8bbcaab1 Fix by Rainer Orth
-
Sergei Golubchik authored
mark auto-inc columns for read/write on INSERT, but only for read on UPDATE
-
Yuchen Pei authored
-
Yuchen Pei authored
When the host is not specified, it defaults to localhost.
-
Yuchen Pei authored
A new column was introduced to the show index output in 10.6 in f691d986 Thus we update the check of the number of columns to be at least 13, rather than exactly 13. Also backport an err number and format from 10.5 for better error messages when the column number is wrong.
-
Rainer Orth authored
There are a large number of uses of `strerror` in the codebase, the local declaration in `storage/connect/tabvct.cpp` is the only one. Given that none is needed elsewhere, I conclude that this instance can simply be removed.
-
- 09 Jan, 2024 2 commits
-
-
Sergei Golubchik authored
initialize THD::rand in THD::init() not in THD::THD(), because the former is also called when a THD is reused - in COM_CHANGE_USER and in taking a THD from the cache. Also use current cycle timer for more unpreditability
-
Alexander Barkov authored
Item_float::neg() did not preserve the "presentation" from "this". So CAST(-1e0 AS UNSIGNED) -- cast from double to unsigned changes its meaning to: CAST(-1 AS UNSIGNED) -- cast signed to undigned Fixing Item_float::neg() to construct the new value for Item_float::presentation as follows: - if the old value starts with minus, then the minus is truncated: '-2e0' -> '2e0' - otherwise, minus sign followed by its old value: '1e0' -> '-1e0'
-
- 08 Jan, 2024 5 commits
-
-
Yuchen Pei authored
The bitmap is temporarily flipped to ~0 for the sake of checking all fields. It needs to be restored because it will be reused in second and subsequent ps execution.
-
Sergei Golubchik authored
Fix by Rainer Orth
-
Sergei Golubchik authored
Fix by Rainer Orth
-
Sergei Golubchik authored
With the result like encryption.innochecksum 'debug' [ skipped ] combination not found instead of *** ERROR: Could not run encryption.innochecksum with 'debug' combination(s)
-
Marko Mäkelä authored
buf_flush_page_cleaner(): A continue or break inside DBUG_EXECUTE_IF actually is a no-op. Use an explicit call to _db_keyword_() to actually avoid advancing the checkpoint. buf_flush_list_now_set(): Invoke os_aio_wait_until_no_pending_writes() to ensure that the page write to the system tablespace is completed.
-
- 04 Jan, 2024 1 commit
-
-
Marko Mäkelä authored
In commit b4ff6456 the signature of mysql_show_var_func was changed, but not all functions of that type were adjusted. When the server is configured with `cmake -DWITH_ASAN=ON` and compiled with clang, runtime errors would be flagged for invoking functions through an incompatible function pointer. Reviewed by: Michael 'Monty' Widenius
-
- 03 Jan, 2024 4 commits
-
-
Igor Babaev authored
If a query contained a CTE whose name coincided with the name of one of the base tables used in the specification of the CTE and the query had at least two references to this CTE in the specifications of other CTEs then processing of the query led to unlimited recursion that ultimately caused a crash of the server. Any secondary non-recursive reference to a CTE requires creation of a copy of the CTE specification. All the references to CTEs in this copy must be resolved. If the specification contains a reference to a base table whose name coincides with the name of then CTE then it should be ensured that this reference in no way can be resolved against the name of the CTE.
-
Marko Mäkelä authored
srv_start(): Move a read only mode startup tweak from innodb_init_params() to the correct location. Also if innodb_force_recovery=6 we will disable the doublewrite buffer, because InnoDB must run in read-only mode to prevent further corruption. This change only affects debug checks. Whenever srv_read_only_mode holds, the buf_pool.flush_list will be empty, that is, there will be no writes of persistent InnoDB data pages. Reviewed by: Thirunarayanan Balathandayuthapani
-
Marko Mäkelä authored
-
Thirunarayanan Balathandayuthapani authored
- innodb.doublewrite_debug should avoid the checkpoint before killing the server. So used debug sync and innodb_flush_sync to avoid the checkpoint completely. Test case allowed to skip on MSAN builder due to extra checkpoint.
-