mariadb

Tree: d93934d7bc

Author	SHA1	Message	Date
Yuchen Pei	98a9c75ea3	MDEV-34659 Use evalp in CREATE SERVER's in init_spider.inc This was already fixed in higher versions.	1 year ago
Yuchen Pei	d3b84ff10d	MDEV-30067 Remove some overly enthusiastic asserts when deleting from a partitioned table When an DDL statement results in a local partition table with partitions not covering all values in the table, a failure is emitted. However, when the table in question is a spider table, the issue does not surface until some future statements (DELETE in the test examples in this commit) are executed. This is consistent with the design of spider which aims to minimise connections with the data node. The resulting error is legitimate and should not result in an assertion failure. Similarly, a partitioned spider table could have misplaced rows, so we remove the other assertion as well.	1 year ago
Yuchen Pei	8a52639ede	MDEV-34716 spider: some trivial cleanups and documentation - document tmp_share, which are temporary spider shares with only one link (no ha) - simplify spider_get_sys_tables_connect_info() where link_idx is always 0	1 year ago
Sergei Golubchik	d5f5062a48	our release packages on rhel8 use bundled readline for historical reasons, but we keep doing it for compatibility	1 year ago
Thirunarayanan Balathandayuthapani	5777d9f282	MDEV-35116 InnoDB fails to set error index for HA_ERR_NULL_IN_SPATIAL - InnoDB fails to set the index information or index number for the spatial index error HA_ERR_NULL_IN_SPATIAL. row_build_spatial_index_key(): Initialize the tmp_mbr array completely. check_if_supported_inplace_alter(): Fix the spelling mistake of alter	1 year ago
Oleksandr Byelkin	a79c9b3812	MDEV-35135 Assertion `!is_cond()' failed in Item_bool_func::val_int / do_select Change val_int with val_bool when it is a condition.	1 year ago
Oleksandr Byelkin	b138f428ea	MDEV-18151 Skipped error returning for GRANT/SET PASSWORD Make message of error not warning.	1 year ago
Oleksandr Byelkin	cc59fbfffa	MDEV-18151 Skipped error returning for GRANT/SET PASSWORD Make error issueing for GRANT and SET PASSWORD the same. Report errors wich were skipped before.	1 year ago
Oleksandr Byelkin	9cd9c3cf16	fix grant5 test to return to the original database.	1 year ago
Marko Mäkelä	cd97caef84	MDEV-33106: Null merge 10.5 into 10.6	1 year ago
Marko Mäkelä	8981ee238a	MDEV-33106 innodb.innodb-lock-inherit-read_commited times out Sometimes, in MariaDB Server 10.5 but apparently not in later branches, the test would hang because con1 and con2 would be blocked in debug_sync (for example, lock_wait_suspend_thread_enter and row_ins_sec_index_entry_dup_locks_created) and therefore blocking the purge of transactions from completing. To prevent an occasional DEBUG_SYNC induced hang in the test, we will wait for everything to be purged, except the last 2 transactions. This change should be null-merged to 10.6, because the test is not failing in 10.6 or later major versions.	1 year ago
Oleksandr Byelkin	1d0e94c55f	Merge branch '10.5' into 10.6	1 year ago
Thirunarayanan Balathandayuthapani	23820f1d79	MDEV-34392 Inplace algorithm violates the foreign key constraint - Fixing the compilation issue for the compiler lesser than gcc-6 Reviewed-by : Marko Mäkelä <marko.makela@mariadb.com>	1 year ago
Sergei Golubchik	4281e0068b	MDEV-35082 HANDLER with FULLTEXT keys is not always rejected need to check for index capabilities also for HANDLER READ NEXT	1 year ago
Sergei Golubchik	23d7dc1ea1	MDEV-35086 Trying to lock mutex when the mutex was already locked (session_tracker.cc), server hangs followup for `9021f40b8e`	1 year ago
Sergei Golubchik	3ea71a2c8e	MDEV-16699 heap-use-after-free in group_concat with compressed or GIS columns Field_blob::store() has special code for GROUP_CONCAT temporary table (to store blob values in Blob_mem_storage - this prevents them from being freed/overwritten when a next row is read). Field_geom and Field_blob_compressed inherit from Field_blob but they have their own ::store() method without this special Blob_mem_storage support. Considering that non-grouping CONCAT() of such fields converts them to plain BLOB, let's do the same for GROUP_CONCAT. To do it, Item_func_group_concat::setup will signal that it's creating a temporary table for GROUP_CONCAT, and Field_blog::make_new_field() override will create base Field_blob when under group concat.	1 year ago
Thirunarayanan Balathandayuthapani	65418ca9ad	MDEV-34392 Inplace algorithm violates the foreign key constraint - Fix the compilation error in gcc-5	1 year ago
Aleksey Midenkov	706a8bcb5b	MDEV-33470 Unique hash index is broken on DML for system-versioned table Hash index is vcol-based wrapper (MDEV-371). row_end is added to unique index. So when row_end is updated unique hash index must be recalculated via vcol_update_fields(). DELETE did not update virtual fields, so DELETE HISTORY was getting wrong hash value. The fix does update_virtual_fields() on vers_update_end() so in every case row_end is updated virtual fields are updated as well.	1 year ago
Aleksey Midenkov	d37bb140b1	MDEV-31297 Create table as select on system versioned tables do not work consistently on replication Row-based replication does not execute CREATE .. SELECT but instead CREATE TABLE. CREATE .. SELECT creates implict system fields on unusual place: in-between declared fields and select fields. That was done because select_field_pos logic requires select fields go last in create_list. So, CREATE .. SELECT on master and CREATE TABLE on slave create system fields on different positions and replication gets field mismatch. To fix this we've changed CREATE .. SELECT to create implicit system fields on usual place in the end and updated select_field_pos for handling this case.	1 year ago
Aleksey Midenkov	4e4c7dd4f5	MDEV-28288 System versioning doesn't support correct work for engine=connect and doesn't always give any warnings/errors Disabled system versioning for connect due to unsupported microseconds (MDEV-15967).	1 year ago
Aleksey Midenkov	5b940bdcfc	MDEV-25060 Freeing overrun buffer, various crashes, ASAN heap-buffer-overflow in _mi_put_key_in_record Rec buffer size depends on vreclength like this: length= MY_MAX(length, info->s->vreclength); The problem is rec buffer is allocated before vreclength is calculated. The fix reallocates rec buffer if vreclength changed. 1. Rec buffer allocated f0 mi_alloc_rec_buff (...) at ../src/storage/myisam/mi_open.c:738 f1 0x00005f4928244516 in mi_open (...) at ../src/storage/myisam/mi_open.c:671 f2 0x00005f4928210b98 in ha_myisam::open (...) at ../src/storage/myisam/ha_myisam.cc:847 f3 0x00005f49273aba41 in handler::ha_open (...) at ../src/sql/handler.cc:3105 f4 0x00005f4927995a65 in open_table_from_share (...) at ../src/sql/table.cc:4320 f5 0x00005f492769f084 in open_table (...) at ../src/sql/sql_base.cc:2024 f6 0x00005f49276a3ea9 in open_and_process_table (...) at ../src/sql/sql_base.cc:3819 f7 0x00005f49276a29b8 in open_tables (...) at ../src/sql/sql_base.cc:4303 f8 0x00005f49276a6f3f in open_and_lock_tables (...) at ../src/sql/sql_base.cc:5250 f9 0x00005f49275162de in open_and_lock_tables (...) at ../src/sql/sql_base.h:509 f10 0x00005f4927a30d7a in open_only_one_table (...) at ../src/sql/sql_admin.cc:412 f11 0x00005f4927a2c0c2 in mysql_admin_table (...) at ../src/sql/sql_admin.cc:603 f12 0x00005f4927a2fda8 in Sql_cmd_optimize_table::execute (...) at ../src/sql/sql_admin.cc:1517 f13 0x00005f49278102e3 in mysql_execute_command (...) at ../src/sql/sql_parse.cc:6180 f14 0x00005f49278012d7 in mysql_parse (...) at ../src/sql/sql_parse.cc:8236 2. vreclength calculated f0 ha_myisam::setup_vcols_for_repair (...) at ../src/storage/myisam/ha_myisam.cc:1002 f1 0x00005f49282138b4 in ha_myisam::optimize (...) at ../src/storage/myisam/ha_myisam.cc:1250 f2 0x00005f49273b4961 in handler::ha_optimize (...) at ../src/sql/handler.cc:4896 f3 0x00005f4927a2d254 in mysql_admin_table (...) at ../src/sql/sql_admin.cc:875 f4 0x00005f4927a2fda8 in Sql_cmd_optimize_table::execute (...) at ../src/sql/sql_admin.cc:1517 f5 0x00005f49278102e3 in mysql_execute_command (...) at ../src/sql/sql_parse.cc:6180 f6 0x00005f49278012d7 in mysql_parse (...) at ../src/sql/sql_parse.cc:8236 FYI backtrace was done with set print frame-info location set print frame-arguments presence set width 80	1 year ago
Alexander Barkov	a931da82fa	MDEV-34123 CONCAT Function Returns Unexpected Empty Set in Query Search conditions were evaluated using val_int(), which was wrong. Fixing the code to use val_bool() instead. Details: - Adding a new item_base_t::IS_COND flag which marks Items used as <search condition> in WHERE, HAVING, JOIN ON, CASE WHEN clauses. The flag is at the parse time. These expressions must be evaluated using val_bool() rather than val_int(). Note, the optimizer creates more Items which are used as search conditions. Most of these items are not marked with IS_COND yet. This is OK for now, but eventually these Items can also be fixed to have the flag. - Adding a method Item::is_cond() which tests if the Item has the IS_COND flag. - Implementing Item_cache_bool. It evaluates the cached expression using val_bool() rather than val_int(). Overriding Type_handler_bool::Item_get_cache() to create Item_cache_bool. - Implementing Item::save_bool_in_field(). It uses val_bool() rather than val_int() to evaluate the expression. - Implementing Type_handler_bool::Item_save_in_field() using Item::save_bool_in_field(). - Fixing all Item_bool_func descendants to implement a virtual val_bool() rather than a virtual val_int(). - To find places where val_int() should be fixed to val_bool(), a few DBUG_ASSERT(!is_cond()) where added into val_int() implementations of selected (most frequent) classes: Item_field Item_str_func Item_datefunc Item_timefunc Item_datetimefunc Item_cache_bool Item_bool_func Item_func_hybrid_field_type Item_basic_constant descendants - Fixing all places where DBUG_ASSERT() happened during an "mtr" run to use val_bool() instead of val_int().	1 year ago
Marko Mäkelä	8a6a4c947a	Cleanup: Replace some is_predefined_tablespace() In some places, there were redundant comparisons against TRX_SYS_SPACE or SRV_TMP_SPACE_ID. The temporary tablespace is never the subject of log-based recovery. Also, consistently check for SRV_SPACE_ID_UPPER_BOUND. Reviewed by: Debarun Barerjee	1 year ago
Marko Mäkelä	b249a059da	MDEV-34850: Busy work while parsing FILE_ records In mariadb-backup --backup, we only have to invoke the undo_space_trunc and log_file_op callbacks as well as validate the mini-transaction checksums. There is absolutely no need to access recv_sys.pages or recv_spaces, or to allocate a decrypt_buf in case of innodb_encrypt_log=ON. This is what the new mode recv_sys_t::store::BACKUP will do. In the skip_the_rest: loop, the main thing is to process all FILE_ records until the end of the log is reached. Additionally, we must process INIT_PAGE and FREE_PAGE records in the same way as they would be during storing == YES. This was measured to reduce the CPU time between the messages "InnoDB: Multi-batch recovery needed at LSN" and "InnoDB: End of log at LSN" by some 20%. recv_sys_t::store: A ternary enumeration that specifies how records should be stored: NO, BACKUP, or YES. recv_sys_t::parse(), recv_sys_t::parse_mtr(), recv_sys_t::parse_pmem(): Replace template<bool store> with template<store storing>. store_freed_or_init_rec(): Simplify some logic. We can look up also the system tablespace. Reviewed by: Debarun Banerjee	1 year ago
Sergei Petrunia	5673cbe094	MDEV-35074: selectivity_notembedded fails with --view-protocol Make the test view-protocol proof: save the contents of optimizer_trace and then we can do many queries against it. Also removed end-of-line spaces.	1 year ago
Yuchen Pei	e9c999caf4	MDEV-34915 Sort output of session track system variable in mysqltest Updated the output format too. Before: -- Tracker : SESSION_TRACK_SYSTEM_VARIABLES -- autocommit -- ON -- time_zone -- SYSTEM -- character_set_client -- latin1 -- character_set_connection -- latin1 -- redirect_url -- After: -- Tracker : SESSION_TRACK_SYSTEM_VARIABLES -- autocommit: ON -- character_set_client: latin1 -- character_set_connection: latin1 -- redirect_url: -- time_zone: SYSTEM	1 year ago
Marko Mäkelä	590718cadc	MDEV-31221 after-merge fix for test The 10.5 test case was accidentally added after 10.11 ones. Fixes up `63913ce5af`	1 year ago
Monty	6f6c1911dc	MDEV-34251 Conditional jump or move depends on uninitialised value in ha_handler_stats::has_stats Fixed by checking handler_stats if it's active instead of thd->variables.log_slow_verbosity & LOG_SLOW_VERBOSITY_ENGINE. Reviewed-by: Sergei Petrunia <sergey@mariadb.com>	1 year ago
Sergei Golubchik	ab15628bbc	MDEV-35050 fix for embedded	1 year ago
Marko Mäkelä	63913ce5af	Merge 10.6 into 10.11	1 year ago
Marko Mäkelä	c6e4ea682c	Merge 10.5 into 10.6	1 year ago
Marko Mäkelä	6878c9d591	MDEV-35050 fixup: ./mtr --embedded	1 year ago
Marko Mäkelä	7e0afb1c73	Merge 10.5 into 10.6	1 year ago
Marko Mäkelä	170e4555e2	MDEV-35040: Fix std::unique deleter for clang++-20 -stdlib=libc++ Thanks to Khem Raj for reporting this.	1 year ago
Marko Mäkelä	23debf214f	MDEV-28091 fixup: Fix another pfs_malloc() stub In commit `0f56e21efa` only one pfs_malloc() stub was fixed to return aligned memory. Also, the MSVC _aligned_malloc() pairs with _aligned_free().	1 year ago
Daniel Black	1f7898f686	mroonga: remove -Wunused-but-set-variable warnings There where unused variable. They were not conditional on defines, so removed them. Added an error handing in proc_object if there was no db as subsequent operations would have failed.	1 year ago
Daniel Black	3723fd1573	MDEV-35007 mroonga should modify source files during build CMake rewriting the tests causes Mroonga to be un-buildable on build environments where there source directory is read only. In the test results, the version wasn't particularly important. Remove the version dependence of tests.	1 year ago
Marko Mäkelä	cc70ca7eab	MDEV-35059 ALTER TABLE...IMPORT TABLESPACE with FULLTEXT SEARCH may corrupt the adaptive hash index build_fts_hidden_table(): Correct a mistake that had been made in commit `903ae30069` (MDEV-30655).	1 year ago
Sergei Petrunia	1cda4726ca	MDEV-34993, part2: backport optimizer_adjust_secondary_key_costs ...and make the fix for MDEV-34993 switchable. It is enabled by default and controlled with @optimizer_adjust_secondary_key_costs=fix_card_multiplier	1 year ago
Sergei Petrunia	8166a5d33d	MDEV-34993: Incorrect cardinality estimation causes poor query plan When calculate_cond_selectivity_for_table() takes into account multi- column selectivities from range access, it tries to take-into account that selectivity for some columns may have been already taken into account. For example, for range access on IDX1 using {kp1, kp2}, the selectivity of restrictions on "kp2" might have already been taken into account to some extent. So, the code tries to "discount" that using rec_per_key[] estimates. This seems to be wrong and unreliable: the "discounting" may produce a rselectivity_multiplier number that hints that the overall selectivity of range access on IDX1 was greater than 1. Do a conservative fix: if we arrive at conclusion that selectivity of range access on condition in IDX1 >1.0, clip it down to 1.	1 year ago
Sergei Golubchik	9021f40b8e	MDEV-35050 Found wrong usage of mutex upon setting plugin session variables	1 year ago
Sergei Golubchik	b1bbdbab9e	cleanup: remove redundant if() likely, a result of auto-merge of two fixes in different versions	1 year ago
Sergei Golubchik	813e592763	compilation failure in CONNECT storage/connect/tabfmt.cpp:419:24: error: '%.3d' directive writing between 3 and 10 bytes into a region of size 5 [-Werror=format-overflow=] 419 \| sprintf(buf, "COL%.3d", i+1);	1 year ago
Sergei Golubchik	5bf543fd43	MDEV-24193 UBSAN: sql/sql_acl.cc:9985:29: runtime error: member access within null pointer of type 'struct TABLE' , ASAN: use-after-poison in handle_grant_table privilege tables do not always have to exist	1 year ago
Sergei Golubchik	2cdcfb644c	MDEV-26314 ST_EQUALS listed twice in information_schema.SQL_FUNCTIONS	1 year ago
Marko Mäkelä	464055fe65	MDEV-34078 Memory leak in InnoDB purge with 32-column PRIMARY KEY row_purge_reset_trx_id(): Reserve large enough offsets for accomodating the maximum width PRIMARY KEY followed by DB_TRX_ID,DB_ROLL_PTR. Reviewed by: Thirunarayanan Balathandayuthapani	1 year ago
Marko Mäkelä	a298dfb84c	MDEV-35053 Crash in purge_sys_t::iterator::free_history_rseg() purge_sys_t::get_page(): Avoid accessing a freed reference to pages[id] after pages.erase(id). This heap-use-after-free would sometimes be caught by AddressSanitizer. purge_sys_t::iterator::free_history_rseg(): Do not crash if undo=nullptr (the database is corrupted). Reviewed by: Debarun Banerjee	1 year ago
Marko Mäkelä	2d031f4a71	MDEV-34973 fixup for POWER,s390x xtest(): Correct the declaration.	1 year ago
Oleksandr Byelkin	8d810e9426	MDEV-29537 Creation of view with UNION and SELECT ... FOR UPDATE in definition is failed with error lock_type is writen in the last SELECT of the unit even if it parsed last, so it should be printed last from the last select of the unit.	1 year ago
Rucha Deodhar	753e7d6d7c	MDEV-27412: JSON_TABLE doesn't properly unquote strings Analysis: The value gets appended as string instead of unescaped json value Fix: Append the value of json in a temporary string and then store it in the field instead of directly storing as string.	1 year ago

1 2 3 4 5 ...

200760 Commits (d93934d7bc51db53c466178cd16e33d5a6c95a78) All Branches Search

200760 Commits (d93934d7bc51db53c466178cd16e33d5a6c95a78)

All Branches