- developers - lists.mariadb.org

Re: [Maria-developers] MDEV-8389 patch review
by Sergey Petrunia 19 Aug '15

19 Aug '15

Hi Diwas, While playing around, I've found and filed: https://mariadb.atlassian.net/browse/MDEV-8649 https://mariadb.atlassian.net/browse/MDEV-8650 I've filed these separately because I think they are not related to MDEV-8389. BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-8389 patch review (part#1)
by Sergey Petrunia 19 Aug '15

19 Aug '15

Hi Diwas, Please find part #1 of the review below. I'll provide more soon. > diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc > index 2998972..a9dc162 100644 > --- a/sql/sql_parse.cc > +++ b/sql/sql_parse.cc > @@ -2466,6 +2466,8 @@ static bool do_execute_sp(THD *thd, sp_head *sp) > lex->first_lists_tables_same(); > /* should be assigned after making first tables same */ > all_tables= lex->query_tables; > + > + check_and_process_table_functions(thd, all_tables); > /* set context for commands which do not use setup_tables */ > select_lex-> > context.resolve_in_table_list_only(select_lex-> > @@ -5581,6 +5583,7 @@ static bool do_execute_sp(THD *thd, sp_head *sp) > } > THD_STAGE_INFO(thd, stage_query_end); > thd->update_stats(); > + //clear_table_functions(thd, all_tables); Please remove the comment. > > goto finish; > > @@ -5697,7 +5700,7 @@ static bool do_execute_sp(THD *thd, sp_head *sp) > thd->mdl_context.release_transactional_locks(); > } > #endif /* WITH_WSREP */ > - > + clear_table_functions(thd, all_tables); > DBUG_RETURN(res || thd->is_error()); > } > > @@ -5709,7 +5712,7 @@ static bool execute_sqlcom_select(THD *thd, TABLE_LIST *all_tables) > bool res; > /* assign global limit variable if limit is not given */ > { > - SELECT_LEX *param= lex->unit.global_parameters(); > + SELECT_LEX *param= lex->unit.global_parameters(); No need for this dummy change. > if (!param->explicit_limit) > param->select_limit= > new (thd->mem_root) Item_int((ulonglong) thd->variables.select_limit); > @@ -9222,3 +9225,182 @@ CHARSET_INFO *find_bin_collation(CHARSET_INFO *cs) > } > return cs; > } > + > +/* > + @brief > + Check for table functions and create corresponding tables to be used in query. > + > + @param thd Thread handle > + @param table_list A list of tables being processed in the query. > + > + @return > + False on success > + @retval > + TRUE on error > +*/ > +bool check_and_process_table_functions(THD *thd, TABLE_LIST *table_list) > +{ > + TABLE_LIST *tables = table_list; > + int err= 0; > + while(tables) > + { > + if (!tables->is_table_function || tables->table_name == NULL) > + { > + tables = tables->next_local; > + continue; > + } > + sp_head *sp; > + sp_name *name; > + LEX_STRING lex_db; > + LEX_STRING lex_name; > + lex_db.length= strlen(tables->db); > + lex_name.length= strlen(tables->table_name); > + lex_db.str= thd->strmake(tables->db, lex_db.length); > + lex_name.str= thd->strmake(tables->table_name, lex_name.length); > + name= new sp_name(lex_db, lex_name, true); > + name->init_qname(thd); > + int err=0; Please switch err from int to bool. > + > + if (!err && (sp = sp_find_routine(thd,TYPE_ENUM_FUNCTION, name, Why assign 'err' and then immediately check it? The code looks redundant. > + &thd->sp_proc_cache, FALSE))) //check if the table is a table-valued function. Please move the comment a few lines down, where the check is actually done. > + { > + if (sp->m_type == TYPE_ENUM_FUNCTION) > + my_error(ER_SP_TABLE_VALUE_EXPECTED, MYF(0)); > + LEX *old_lex= thd->lex, newlex; > + thd->lex= &newlex; > + lex_start(thd); > + ulonglong saved_mode= thd->variables.sql_mode; > + Table_specification_st create_info; > + Alter_info alter_info; > + create_info.init(); > + alter_info.reset(); > + create_info.default_table_charset= NULL; > + TABLE_LIST *create_table; > + create_info.options|= HA_LEX_CREATE_TMP_TABLE; > + LEX_STRING tablename,dbname; > + if (strlen(sp->m_table_alias.str) > 0) { > + tablename.str= sp->m_table_alias.str; > + tablename.length= strlen(sp->m_table_alias.str); > + } sp->m_table_alias is the alias of the return table of the table function. When can it have zero length? Please add a comment about this here, and also add a testcase in the .test file > + else { > + tablename.str= tables->table_name; > + tablename.length= strlen(tables->table_name); > + } > + dbname.str= tables->db; > + dbname.length= strlen(tables->db); > + Table_ident *new_ident= new Table_ident(thd, dbname, tablename, 0); > + if (!thd->lex->select_lex.add_table_to_list(thd, new_ident, NULL, > + TL_OPTION_UPDATING, > + TL_WRITE, MDL_EXCLUSIVE)) > + err= 1; > + create_info.use_default_db_type(thd); > + thd->lex->query_tables->open_strategy= TABLE_LIST::OPEN_STUB; > + thd->lex->create_info.default_table_charset= NULL; > + thd->lex->name= null_lex_str; > + create_table= thd->lex->create_last_non_select_table= thd->lex->last_table(); > + thd->lex->current_select= &thd->lex->select_lex; > + create_table_set_open_action_and_adjust_tables(thd->lex); > + alter_info.create_list= sp->m_cols_list; > + err= mysql_create_table(thd, create_table, &create_info, &alter_info); Ok, so now it is using mysql_create_table instead of parsing the table definition. This is a step forward. However, I still think that mysql_create_table() does too much. It is a function that used for regular CREATE TABLE statements. When I trace a query, I still see that the created temporary table uses InnoDB storage engine. I think, more effort is needed to make temporary table creation work in a more similar to how other parts of code create temporary tables. I think I have also found a problem with how name resolution process works. I'll elaborate in a separate email. > + thd->variables.option_bits|= OPTION_KEEP_LOG; What is this for? please add a comment. > + if (!err) > + close_thread_tables(thd); > + execute_table_function(thd,sp,tables->table_func_params); > + thd->variables.sql_mode= saved_mode; > + lex_end(thd->lex); > + thd->lex= old_lex; > + if (strlen(sp->m_table_alias.str) > 0) { > + tables->table_name= sp->m_table_alias.str; > + tables->mdl_request.init(tables->mdl_request.key.mdl_namespace(), > + tables->db,sp->m_table_alias.str, > + tables->mdl_request.type,tables->mdl_request.duration); > + } What does the above statement do? Please add a comment. > + } > + tables = tables->next_local; > + } > + close_thread_tables(thd); > + return err; > +} ... > +bool execute_table_function(THD *thd, sp_head *sp, List<Item> *value_list) > +{ > + /* bits that should be cleared in thd->server_status */ > + uint bits_to_be_cleared= 0; > + if (sp->m_flags & sp_head::MULTI_RESULTS) > + { > + if (!(thd->client_capabilities & CLIENT_MULTI_RESULTS)) > + { > + /* The client does not support multiple result sets being sent back */ > + my_error(ER_SP_BADSELECT, MYF(0), sp->m_qname.str); > + return 1; > + } > + /* > + If SERVER_MORE_RESULTS_EXISTS is not set, > + then remember that it should be cleared > + */ > + bits_to_be_cleared= (~thd->server_status & > + SERVER_MORE_RESULTS_EXISTS); > + thd->server_status|= SERVER_MORE_RESULTS_EXISTS; > + } Does the above make any sense here? This is a piece of logic from code that executes stored *procedures*. A stored procedure may have multiple SELECT statements. Executing this stored procedure with "CALL sp(...)" may produce multiple result sets. So this logic checks that the client is able to receive multiple resultsets. A stored *function* may not emit resultsets. It must return a scalar value, in our case it must return a table. > + ha_rows select_limit= thd->variables.select_limit; > + thd->variables.select_limit= HA_POS_ERROR; > + > + /* > + We never write CALL statements into binlog: > + - If the mode is non-prelocked, each statement will be logged > + separately. > + - If the mode is prelocked, the invoking statement will care > + about writing into binlog. > + So just execute the statement. > + */ > + int res= sp->execute_procedure(thd, value_list); > + > + thd->variables.select_limit= select_limit; > + thd->server_status&= ~bits_to_be_cleared; > + > + if (res) > + { > + DBUG_ASSERT(thd->is_error() || thd->killed); > + return 1; // Substatement should already have sent error > + } > + > + my_ok(thd, (thd->get_row_count_func() < 0) ? 0 : thd->get_row_count_func()); What is the above for? As far as I understand, the code in this function was based on do_execute_sp(). Which is not a problem, however, do_execute_sp() is used to handle "CALL stored_proc" or compound statements, which should produce "Ok" if they have not produced a resultset. There is no such need when using a table function. Upper query will produce a resultset which it will return to the client. > + thd->reset_query_timer(); What is this for? > + thd->lex->unit.cleanup(); > + thd->mdl_context.release_transactional_locks(); > + thd->reset_for_next_command(); > + return 0; > +} > \ No newline at end of file ... > diff --git a/sql/table.h b/sql/table.h > index 622a3b2..5eddd3d 100644 > --- a/sql/table.h > +++ b/sql/table.h > @@ -2010,6 +2010,7 @@ struct TABLE_LIST > List<Item> used_items; > /* Sublist (tail) of persistent used_items */ > List<Item> persistent_used_items; > + List<Item> *table_func_params; //parameters used for table function. > Item **materialized_items; > > /* View creation context. */ > @@ -2059,7 +2060,7 @@ struct TABLE_LIST > from I_S should be done by create_sort_index() or by JOIN::exec.) > */ > enum enum_schema_table_state schema_table_state; > - > + bool is_table_function; //Check whether the table is actually a table function. Do I get it right that a TABLE_LIST object that denotes a Table Function invocation always has table_func_params != NULL? If yes, please remove 'bool is_table_function' and add a 'bool is_table_function()' function instead which checks for table_func_params!=NULL. > /* Something like a "query plan" for reading INFORMATION_SCHEMA table */ > IS_table_read_plan *is_table_read_plan; > > @@ -2277,6 +2278,18 @@ struct TABLE_LIST > return false; > } > void set_lock_type(THD* thd, enum thr_lock_type lock); > + void fill_table_func_params(List<Item> param_list) > + { > + table_func_params= new List<Item>; > + List_iterator_fast<Item> it(param_list); > + Item *param; > + param= it++; > + while(param) > + { > + table_func_params->push_back(param); > + param= it++; > + } > + } > > private: > bool prep_check_option(THD *thd, uint8 check_opt_type); BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-7502 Provisioning of slave - review
by Kristian Nielsen 17 Aug '15

17 Aug '15

Martin Kaluznik <martin.kaluznik(a)gmail.com> writes: > Earlier you mentioned, that you are able to get my code on build > server to try compile and test runs on various platforms, it would be Ok. I have added you as a developer on https://github.com/MariaDB/server. The way it works is that you push your changes to that repository in a branch that starts with "bb-", eg. "bb-10.1-f4rnham". Then Buildbot will pick up your changes and start building (can take some hours or even a day or two in busy periods). Your builds will appear then on this page: https://buildbot.askmonty.org/buildbot/grid?branch=bb-10.1-f4rnham (Please be careful about not accidentally pushing to main branches (or any other branches) in the main repository until we've completed the review process.) Then when the builds are done, you can use the web interface to check for any failures on special platforms or test configuration, and fix them as appropriate. Meanwhile, I will read through all of your patch, and get back to you with comments. > dbug.c,my_dbug.h > When line like SET GLOBAL debug_dbug="+d,flag"; is executed in > test case, it acquires THR_LOCK_dbug, but when it is tested for > (DBUG_EXECUTE_IF) no lock is used - and it causes data races in > replication threads Hm. The way it usually works is that the test cases SET GLOBAL debug_dbug while the slave threads are stopped. That way there should be no races? You might be right that there is a need for a lock, but I would prefer if that was made as a separate change from MDEV-7502, if possible. But it can be tricky to properly control the slave threads for testing, I will keep this point in mind when I check your test cases, and see if I can come up with concrete suggestions where needed. > Edited *.test,*.result > Updated old test cases, most of them because I have added new log > event type, what increased length of format description event and all > positions in test cases have shifted Right, annoying but expected. The Buildbot might catch a few more that are missed by a standard test run. - Kristian.

3 2

[Maria-developers] Unexpected contents of mysql.gtid_slave_pos
by Giuseppe Maxia 13 Aug '15

13 Aug '15

Using MariaDB 10.0.20, when I set a multi-master environment in which one of the masters has log-slave-updated enabled, I get this strange behavior: -- node#1 select @@server_id; +-------------+ | @@server_id | +-------------+ | 101 | +-------------+ use test; create table t1 (i int not null primary key, msg varchar(50), d date, t time, dt datetime, ts timestamp); use mysql; select * from gtid_slave_pos; +-----------+--------+-----------+--------+ | domain_id | sub_id | server_id | seq_no | +-----------+--------+-----------+--------+ | 0 | 1 | 101 | 1 | +-----------+--------+-----------+--------+ Why is this happening? Is there a reason for the server to log its own transactions? The transaction originated from this same server, but it is effectively filtered off when it comes back from the other node (otherwise we would have a conflict). Note: I get the same behavior if I use multi-source setup where one of the nodes has log-slave-updates enabled. Using different domains for each master does not affect the outcome. Thanks Giuseppe

2 4

[Maria-developers] Interested in Contributing Code to Maria DB.
by Alangi Derick 08 Aug '15

08 Aug '15

Hi, I am Alangi Derick and a computer engineering student(specialty: software engineering) from the University of Buea. I am a junior and i code in C. I have found interest in Maria DB. I have used many DBMS but i will like to develop for MariaDB since it is Open source and it is the one that has caught my interest. I have some contributions to make to this organisation in terms of the Maria DB project. I am very proficient in my skills of programming and will first of all start by fixing some bugs in the Maria DB code base before writing a whole module(implementation of my idea) to add to maria db. I will like if the entire community welcomes and support me so as to make Maria DB a better DBMS for everyone. Thanks and hope to participate in the development. Regards Alangi Derick Ndimnain

2 1

Re: [Maria-developers] [Commits] 686761f: MDEV-6066: Merge new defaults from 5.6 and 5.7 (part 1 (no hash serch) v2)
by Sergei Golubchik 07 Aug '15

07 Aug '15

Hi, Sanja! Summary: The implementation looks very much ok, all comments below are about details, nothing big. Well done! On Jul 14, sanja(a)mariadb.com wrote: > revision-id: 686761fbb0fc36e4fa781b68aba24af21599822d > parent(s): 302bf7c4664b904482ecc133476e822d497b114d > committer: Oleksandr Byelkin > branch nick: server could you update your post-commit trigger? thanks. > timestamp: 2015-07-14 00:11:25 +0200 > message: > > MDEV-6066: Merge new defaults from 5.6 and 5.7 (part 1 (no hash serch) v2) > diff --git a/mysql-test/extra/binlog_tests/mysqlbinlog_row_engine.inc b/mysql-test/extra/binlog_tests/mysqlbinlog_row_engine.inc > index 95440ab..c072d50 100644 > --- a/mysql-test/extra/binlog_tests/mysqlbinlog_row_engine.inc > +++ b/mysql-test/extra/binlog_tests/mysqlbinlog_row_engine.inc > @@ -16,6 +16,7 @@ > # > > --source include/have_log_bin.inc > +set sql_mode=""; please split this big commit in 'new defaults' and 'autoset' commits. > > SET NAMES 'utf8'; > #SHOW VARIABLES LIKE 'character_set%'; > diff --git a/mysql-test/r/analyze_stmt_privileges.result b/mysql-test/r/analyze_stmt_privileges.result > index d382b0b..05a35e7 100644 > --- a/mysql-test/r/analyze_stmt_privileges.result > +++ b/mysql-test/r/analyze_stmt_privileges.result > @@ -7,7 +7,7 @@ use db; > create table t1 (i int, c varchar(8)); > insert into t1 values (1,'foo'),(2,'bar'),(3,'baz'),(4,'qux'); > create view v1 as select * from t1 where i > 1; > -grant ALL on db.v1 to u1@localhost; > +set statement sql_mode="" for grant ALL on db.v1 to u1@localhost; heh. I think it would've been less confusing to put here an explicit CREATE USER statement > connect con1,localhost,u1,,; > select * from db.t1; > ERROR 42000: SELECT command denied to user 'u1'@'localhost' for table 't1' > diff --git a/mysql-test/r/blackhole_plugin.result b/mysql-test/r/blackhole_plugin.result > index 4ef9fa0..ab106f8 100644 > --- a/mysql-test/r/blackhole_plugin.result > +++ b/mysql-test/r/blackhole_plugin.result > @@ -1,7 +1,9 @@ > +set sql_mode=""; you could've put --error 1286 instead. > CREATE TABLE t1(a int) ENGINE=BLACKHOLE; > Warnings: > Warning 1286 Unknown storage engine 'BLACKHOLE' > Warning 1266 Using storage engine MyISAM for table 't1' > +set sql_mode=default; > DROP TABLE t1; > INSTALL PLUGIN blackhole SONAME 'ha_blackhole.so'; > INSTALL PLUGIN BLACKHOLE SONAME 'ha_blackhole.so'; > diff --git a/mysql-test/r/change_user.result b/mysql-test/r/change_user.result > index 18c53a5..bd87398 100644 > --- a/mysql-test/r/change_user.result > +++ b/mysql-test/r/change_user.result > @@ -1,3 +1,5 @@ > +set sql_mode=""; > +set global secure_auth=0; > grant select on test.* to test_nopw; > grant select on test.* to test_oldpw identified by password "09301740536db389"; > grant select on test.* to test_newpw identified by "newpw"; you could've used "create user" instead of "grant select on test.* to" and that would work without resetting sql_mode > diff --git a/mysql-test/r/concurrent_innodb_safelog.result b/mysql-test/r/concurrent_innodb_safelog.result > index 24a84af..03d20f4 100644 > --- a/mysql-test/r/concurrent_innodb_safelog.result > +++ b/mysql-test/r/concurrent_innodb_safelog.result > @@ -1,4 +1,5 @@ > SET GLOBAL TRANSACTION ISOLATION LEVEL REPEATABLE READ; > +SET SQL_MODE=""; same. basically I'd suggest to fix tests instead of setting sql_mode="". i suspect it's mostly adding create user or replacing grant with create user. sometimes it's about removing engine=xxx and you're doing that already. > SELECT @@global.tx_isolation; > @@global.tx_isolation > REPEATABLE-READ > diff --git a/mysql-test/r/func_compress.result b/mysql-test/r/func_compress.result > index 9fde006..4763d87 100644 > --- a/mysql-test/r/func_compress.result > +++ b/mysql-test/r/func_compress.result > @@ -1,3 +1,4 @@ > +set global max_allowed_packet=1048576; again, I'd rather fix the test > select @test_compress_string:='string for test compress function aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa '; > @test_compress_string:='string for test compress function aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa ' > string for test compress function aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa > diff --git a/mysql-test/r/host_cache_size_functionality.result b/mysql-test/r/host_cache_size_functionality.result > index e7f9e09..ffa8d71 100644 > --- a/mysql-test/r/host_cache_size_functionality.result > +++ b/mysql-test/r/host_cache_size_functionality.result > @@ -9,7 +9,7 @@ SELECT COUNT(@@GLOBAL.Host_Cache_Size) > set @Default_host_cache_size=128; > select @@global.Host_Cache_Size=@Default_host_cache_size; > @@global.Host_Cache_Size=@Default_host_cache_size > -1 > +0 > 1 Expected Now that looks weird. > '#---------------------WL6372_VAR_6_02----------------------#' > # Restart server with Host_Cache_Size 1 > diff --git a/mysql-test/r/index_intersect.result b/mysql-test/r/index_intersect.result > index 1337c3f..edc52db 100644 > --- a/mysql-test/r/index_intersect.result > +++ b/mysql-test/r/index_intersect.result > @@ -1035,7 +1032,7 @@ EXPLAIN > SELECT * FROM t1 > WHERE (f1 < 535 OR f1 > 985) AND ( f4='r' OR f4 LIKE 'a%' ) ; > id select_type table type possible_keys key key_len ref rows Extra > -1 SIMPLE t1 range PRIMARY,f4 f4 35 NULL # Using index condition; Using where > +1 SIMPLE t1 index_merge PRIMARY,f4 PRIMARY,f4 4,39 NULL # Using sort_intersect(PRIMARY,f4); Using where why? > SELECT * FROM t1 > WHERE (f1 < 535 OR f1 > 985) AND ( f4='r' OR f4 LIKE 'a%' ) ; > f1 f4 f5 > diff --git a/mysql-test/r/mysqld--help.result b/mysql-test/r/mysqld--help.result > index fc22af6..1f2a8cf 100644 > --- a/mysql-test/r/mysqld--help.result > +++ b/mysql-test/r/mysqld--help.result > @@ -890,6 +890,7 @@ The following options may be given as the first argument: > write privileges to the mysql.user table. > --secure-auth Disallow authentication for accounts that have old > (pre-4.1) passwords > + (Defaults to on; use --skip-secure-auth to disable.) You know what, it would be nice if the user would know what options support the --autoset prefix. I think here's a good place for it, the help for back_log can end with gets very many connection requests in a very short time (Automatically configured unless set explicitly) Of course, this line should be automatically generated by my_print_help(), not manually added to the help text of every autoset variable. > --secure-file-priv=name > Limit LOAD DATA, SELECT ... OUTFILE, and LOAD_FILE() to > files within specified directory > diff --git a/mysql-test/r/selectivity.result b/mysql-test/r/selectivity.result > index 620bdc6..4238359 100644 > --- a/mysql-test/r/selectivity.result > +++ b/mysql-test/r/selectivity.result > @@ -1298,10 +1298,10 @@ Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b`,`test`.`t2`.`c` A > select * from t1, t2, t1 as t3 > where t1.b=t2.c and t2.d=t3.a and t3.b<5 and t1.a < 2000; > a b c d a b > -1495 89366 89366 28296 28296 3 > -961 24512 24512 85239 85239 4 > 1063 89366 89366 28296 28296 3 > +1495 89366 89366 28296 28296 3 > 221 56120 56120 28296 28296 3 > +961 24512 24512 85239 85239 4 why? > set optimizer_use_condition_selectivity=3; > explain extended > select * from t1, t2, t1 as t3 > @@ -1315,10 +1315,10 @@ Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b`,`test`.`t2`.`c` A > select * from t1, t2, t1 as t3 > where t1.b=t2.c and t2.d=t3.a and t3.b<5 and t1.a < 2000; > a b c d a b > -961 24512 24512 85239 85239 4 > -1495 89366 89366 28296 28296 3 > 1063 89366 89366 28296 28296 3 > +1495 89366 89366 28296 28296 3 > 221 56120 56120 28296 28296 3 > +961 24512 24512 85239 85239 4 why? > set optimizer_use_condition_selectivity=@save_optimizer_use_condition_selectivity; > drop table t1,t2; > set histogram_type=@save_histogram_type; > diff --git a/mysql-test/r/selectivity_innodb.result b/mysql-test/r/selectivity_innodb.result > index 0acbb46..daf2807 100644 > --- a/mysql-test/r/selectivity_innodb.result > +++ b/mysql-test/r/selectivity_innodb.result > @@ -1308,10 +1308,10 @@ Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b`,`test`.`t2`.`c` A > select * from t1, t2, t1 as t3 > where t1.b=t2.c and t2.d=t3.a and t3.b<5 and t1.a < 2000; > a b c d a b > -1495 89366 89366 28296 28296 3 > -961 24512 24512 85239 85239 4 > 1063 89366 89366 28296 28296 3 > +1495 89366 89366 28296 28296 3 > 221 56120 56120 28296 28296 3 > +961 24512 24512 85239 85239 4 why? > set optimizer_use_condition_selectivity=3; > explain extended > select * from t1, t2, t1 as t3 > @@ -1325,10 +1325,10 @@ Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b`,`test`.`t2`.`c` A > select * from t1, t2, t1 as t3 > where t1.b=t2.c and t2.d=t3.a and t3.b<5 and t1.a < 2000; > a b c d a b > -961 24512 24512 85239 85239 4 > -1495 89366 89366 28296 28296 3 > 1063 89366 89366 28296 28296 3 > +1495 89366 89366 28296 28296 3 > 221 56120 56120 28296 28296 3 > +961 24512 24512 85239 85239 4 why? > set optimizer_use_condition_selectivity=@save_optimizer_use_condition_selectivity; > drop table t1,t2; > set histogram_type=@save_histogram_type; > diff --git a/mysql-test/suite/funcs_1/r/is_columns_is_embedded.result b/mysql-test/suite/funcs_1/r/is_columns_is_embedded.result > index c314165..bd49492 100644 > --- a/mysql-test/suite/funcs_1/r/is_columns_is_embedded.result > +++ b/mysql-test/suite/funcs_1/r/is_columns_is_embedded.result > @@ -167,9 +167,9 @@ def information_schema GEOMETRY_COLUMNS MAX_PPR 12 0 NO tinyint NULL NULL 3 0 NU > def information_schema GEOMETRY_COLUMNS SRID 13 0 NO smallint NULL NULL 5 0 NULL NULL NULL smallint(5) > def information_schema GEOMETRY_COLUMNS STORAGE_TYPE 9 0 NO tinyint NULL NULL 3 0 NULL NULL NULL tinyint(2) > def information_schema GLOBAL_STATUS VARIABLE_NAME 1 NO varchar 64 192 NULL NULL NULL utf8 utf8_general_ci varchar(64) > -def information_schema GLOBAL_STATUS VARIABLE_VALUE 2 NO varchar 1024 3072 NULL NULL NULL utf8 utf8_general_ci varchar(1024) > +def information_schema GLOBAL_STATUS VARIABLE_VALUE 2 NO varchar 2048 6144 NULL NULL NULL utf8 utf8_general_ci varchar(2048) > def information_schema GLOBAL_VARIABLES VARIABLE_NAME 1 NO varchar 64 192 NULL NULL NULL utf8 utf8_general_ci varchar(64) > -def information_schema GLOBAL_VARIABLES VARIABLE_VALUE 2 NO varchar 1024 3072 NULL NULL NULL utf8 utf8_general_ci varchar(1024) > +def information_schema GLOBAL_VARIABLES VARIABLE_VALUE 2 NO varchar 2048 6144 NULL NULL NULL utf8 utf8_general_ci varchar(2048) why? > def information_schema INDEX_STATISTICS INDEX_NAME 3 NO varchar 192 576 NULL NULL NULL utf8 utf8_general_ci varchar(192) > def information_schema INDEX_STATISTICS ROWS_READ 4 0 NO bigint NULL NULL 19 0 NULL NULL NULL bigint(21) > def information_schema INDEX_STATISTICS TABLE_NAME 2 NO varchar 192 576 NULL NULL NULL utf8 utf8_general_ci varchar(192) > diff --git a/mysql-test/suite/perfschema/r/stage_mdl_global.result b/mysql-test/suite/perfschema/r/stage_mdl_global.result > index 1a6f51a..de5df8f 100644 > --- a/mysql-test/suite/perfschema/r/stage_mdl_global.result > +++ b/mysql-test/suite/perfschema/r/stage_mdl_global.result > @@ -6,6 +6,8 @@ user1 statement/sql/flush flush tables with read lock > username event_name nesting_event_type > username event_name nesting_event_type > user1 stage/sql/init STATEMENT > +user1 stage/sql/Waiting for query cache lock STATEMENT > +user1 stage/sql/init STATEMENT why? > user1 stage/sql/query end STATEMENT > user1 stage/sql/closing tables STATEMENT > user1 stage/sql/freeing items STATEMENT > diff --git a/mysys/my_getopt.c b/mysys/my_getopt.c > index 88fa3d4..1d32833 100644 > --- a/mysys/my_getopt.c > +++ b/mysys/my_getopt.c > @@ -47,14 +47,15 @@ static char *check_struct_option(char *cur_arg, char *key_name); > order of their arguments must correspond to each other. > */ > static const char *special_opt_prefix[]= > -{"skip", "disable", "enable", "maximum", "loose", 0}; > +{"skip", "disable", "enable", "maximum", "loose", "autoset", 0}; So, I'm not very happy with the --autoset- prefix. Because it cannot be used from the SQL. In SQL we have to use SET variable=AUTO; or something like that. And it would be good to use the same approach on the command line and from SQL. On the other hand, =AUTO has its problems too - it conflicts with the AUTO value of the enum/set variables. I don't have a good answer to this :( > static const uint special_opt_prefix_lengths[]= > -{ 4, 7, 6, 7, 5, 0}; > +{ 4, 7, 6, 7, 5, 7, 0}; > enum enum_special_opt > -{ OPT_SKIP, OPT_DISABLE, OPT_ENABLE, OPT_MAXIMUM, OPT_LOOSE}; > +{ OPT_SKIP, OPT_DISABLE, OPT_ENABLE, OPT_MAXIMUM, OPT_LOOSE, OPT_AUTOSET}; > > char *disabled_my_option= (char*) "0"; > char *enabled_my_option= (char*) "1"; > +char *autoset_my_option= (char*) ""; > > /* > This is a flag that can be set in client programs. 0 means that > @@ -458,6 +462,36 @@ int handle_options(int *argc, char ***argv, > } > argument= optend; > } > + else if (option_is_autoset) > + { > + if (optend) > + { > + my_getopt_error_reporter(ERROR_LEVEL, > + "%s: automatic setup request of " > + "option '--%s' cannot take an argument", > + my_progname, optp->name); heh, another argument agains --autoset- prefix. With the =AUTO value one cannot possibly have this error. So, perhaps, I'm *slighty* leaning towards the =AUTO vs --autoset. > + > + DBUG_RETURN(EXIT_NO_ARGUMENT_ALLOWED); > + } > + /* > + We support automatic setup only via get_one_option and only for > + marked options. > + */ > + if (!get_one_option || > + !(optp->var_type & GET_AUTO)) > + { > + my_getopt_error_reporter(option_is_loose ? > + WARNING_LEVEL : ERROR_LEVEL, > + "%s: automatic setup request is " > + "unsupported by option '--%s'", > + my_progname, optp->name); > + if (!option_is_loose) > + return EXIT_ARGUMENT_INVALID; > + continue; > + } > + else > + argument= autoset_my_option; > + } > else if (optp->arg_type == REQUIRED_ARG && !optend) > { > /* Check if there are more arguments after this one, > diff --git a/sql/set_var.cc b/sql/set_var.cc > index c65ca3d..2251c36 100644 > --- a/sql/set_var.cc > +++ b/sql/set_var.cc > @@ -1075,7 +1076,7 @@ int fill_sysvars(THD *thd, TABLE_LIST *tables, COND *cond) > // NUMERIC_MAX_VALUE > // NUMERIC_BLOCK_SIZE > bool is_unsigned= true; > - switch (var->option.var_type) > + switch (vartype) ah, yes. thanks. > { > case GET_INT: > case GET_LONG: > @@ -1176,3 +1177,20 @@ void mark_sys_var_value_origin(void *ptr, enum sys_var::where here) > DBUG_ASSERT(found); // variable must have been found > } > > +enum sys_var::where get_sys_var_value_origin(void *ptr) > +{ > + DBUG_ASSERT(!mysqld_server_started); // only to be used during startup > + > + for (uint i= 0; i < system_variable_hash.records; i++) > + { > + sys_var *var= (sys_var*) my_hash_element(&system_variable_hash, i); > + if (var->option.value == ptr) > + { > + return var->value_origin; //first match > + } > + } > + > + DBUG_ASSERT(1); // variable must have been found you mean DBUG_ASSERT(0); here > + return sys_var::CONFIG; > +} > + > diff --git a/sql/mysqld.cc b/sql/mysqld.cc > index 43a3f0e..21a1198 100644 > --- a/sql/mysqld.cc > +++ b/sql/mysqld.cc > @@ -4316,6 +4316,27 @@ static int init_common_variables() > } > #endif /* HAVE_SOLARIS_LARGE_PAGES */ > > + > + /* Fix host_cache_size. */ > + if (IS_SYSVAR_AUTOSIZE(&host_cache_size)) > + { > + if (max_connections <= 628 - 128) > + SYSVAR_AUTOSIZE(host_cache_size, 128 + max_connections); > + else if (max_connections <= ((ulong)(2000 - 628)) * 20 + 500) > + SYSVAR_AUTOSIZE(host_cache_size, 628 + ((max_connections - 500) / 20)); > + else > + SYSVAR_AUTOSIZE(host_cache_size, 2000); > + } > + > + /* Fix back_log */ > + if (back_log == 0 || IS_SYSVAR_AUTOSIZE(&back_log)) why back_log==0 ? > + { > + if ((900 - 50) * 5 >= max_connections) > + SYSVAR_AUTOSIZE(back_log, (50 + max_connections / 5)); > + else > + SYSVAR_AUTOSIZE(back_log, 900); > + } > + > /* connections and databases needs lots of files */ > { > uint files, wanted_files, max_open_files; > diff --git a/sql/sys_vars.ic b/sql/sys_vars.ic > index 7c0359a..b04bdb5 100644 > --- a/sql/sys_vars.ic > +++ b/sql/sys_vars.ic > @@ -226,6 +227,7 @@ class Sys_var_integer: public sys_var > typedef Sys_var_integer<int, GET_INT, SHOW_SINT> Sys_var_int; > typedef Sys_var_integer<uint, GET_UINT, SHOW_UINT> Sys_var_uint; > typedef Sys_var_integer<ulong, GET_ULONG, SHOW_ULONG> Sys_var_ulong; > +typedef Sys_var_integer<ulong, (GET_ULONG|GET_AUTO), SHOW_ULONG> Sys_var_aulong; Hmm, why did you make it a new type? I think the *type* of the variable doesn't depend on the auto-set behavior. The type defines storage size, how to print the value, etc. While auto-set is just a behavior modifier. I'd rather put it in flags: - READ_ONLY GLOBAL_VAR(back_log), CMD_LINE(REQUIRED_ARG), + AUTO_SET READ_ONLY GLOBAL_VAR(back_log), CMD_LINE(REQUIRED_ARG), > typedef Sys_var_integer<ha_rows, GET_HA_ROWS, SHOW_HA_ROWS> Sys_var_harows; > typedef Sys_var_integer<ulonglong, GET_ULL, SHOW_ULONGLONG> Sys_var_ulonglong; > typedef Sys_var_integer<long, GET_LONG, SHOW_SLONG> Sys_var_long; > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index 7a7b9d1..6d90305 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -360,12 +360,12 @@ static Sys_var_mybool Sys_automatic_sp_privileges( > GLOBAL_VAR(sp_automatic_privileges), > CMD_LINE(OPT_ARG), DEFAULT(TRUE)); > > -static Sys_var_ulong Sys_back_log( > +static Sys_var_aulong Sys_back_log( > "back_log", "The number of outstanding connection requests " > "MySQL can have. This comes into play when the main MySQL thread " s/MySQL/MariaDB/ I suppose > "gets very many connection requests in a very short time", > READ_ONLY GLOBAL_VAR(back_log), CMD_LINE(REQUIRED_ARG), > - VALID_RANGE(1, 65535), DEFAULT(150), BLOCK_SIZE(1)); > + VALID_RANGE(0, 65535), DEFAULT(150), BLOCK_SIZE(1)); > > static Sys_var_charptr Sys_basedir( > "basedir", "Path to installation directory. All paths are " > @@ -2751,16 +2743,38 @@ static bool check_query_cache_type(sys_var *self, THD *thd, set_var *var) > my_error(ER_QUERY_CACHE_IS_DISABLED, MYF(0)); > return true; > } > - if (var->type != OPT_GLOBAL && > - global_system_variables.query_cache_type == 0 && > - var->value->val_int() != 0) > + > + if (var->type != OPT_GLOBAL && global_system_variables.query_cache_type == 0) > { > - my_error(ER_QUERY_CACHE_IS_GLOBALY_DISABLED, MYF(0)); > - return true; > - } > + uint value= 0 /*default is off*/; > + if (var->value) > + { > + if (var->value->result_type() == INT_RESULT) > + value= var->value->val_int(); > + else > + { > + char buff[STRING_BUFFER_USUAL_SIZE]; > + String str(buff, sizeof(buff), system_charset_info), *res; > + if (!(res=var->value->val_str(&str))) > + return true; > + if (res->length() != 3 || > + my_toupper(res->charset(), res->ptr()[0]) != 'O' || > + my_toupper(res->charset(), res->ptr()[1]) != 'F' || > + my_toupper(res->charset(), res->ptr()[2]) != 'F') eh? query_cache_type is ENUM variable, there is no need to compare string values manually. > + value= 1; // set to something not 0 > + } > + } > > + if (value != 0) > + { > + my_error(ER_QUERY_CACHE_IS_GLOBALY_DISABLED, MYF(0)); > + return true; > + } > + } > return false; > } > + > + > static bool fix_query_cache_type(sys_var *self, THD *thd, enum_var_type type) > { > if (type != OPT_GLOBAL) > diff --git a/unittest/mysys/CMakeLists.txt b/unittest/mysys/CMakeLists.txt > index 8c09a46..6564e09 100644 > --- a/unittest/mysys/CMakeLists.txt > +++ b/unittest/mysys/CMakeLists.txt > @@ -13,7 +13,7 @@ > # along with this program; if not, write to the Free Software > # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA > > -MY_ADD_TESTS(bitmap base64 my_vsnprintf my_atomic my_rdtsc lf my_malloc > +MY_ADD_TESTS(bitmap base64 my_vsnprintf my_atomic my_rdtsc lf my_malloc my_getopt Ah, that's great, thanks! Long overdue. > LINK_LIBRARIES mysys) > > MY_ADD_TESTS(ma_dyncol Regards, Sergei

2 2

Re: [Maria-developers] [Commits] 69704b6: MDEV-6887 Optimize sys_vars.sysvar_* tests
by Sergei Golubchik 06 Aug '15

06 Aug '15

Hi, Elena! On Aug 06, elenst(a)montyprogram.com wrote: > revision-id: 69704b6c6f895fa706f83bcd3bcefe267aaf8e57 > parent(s): afd59b575a75ebbc57f71ce2865fdff85e3e233b > committer: Elena Stepanova > branch nick: bb-10.1-elenst > timestamp: 2015-08-06 02:55:43 +0300 > message: > > MDEV-6887 Optimize sys_vars.sysvar_* tests > > diff --git a/mysql-test/suite/sys_vars/inc/sysvars_server.inc b/mysql-test/suite/sys_vars/inc/sysvars_server.inc > index cb06b40..98bffd4 100644 > --- a/mysql-test/suite/sys_vars/inc/sysvars_server.inc > +++ b/mysql-test/suite/sys_vars/inc/sysvars_server.inc > @@ -11,8 +12,9 @@ set sql_mode=ansi_quotes; > # global_value_origin=SQL > set global div_precision_increment=5; > > ---replace_regex /^\/\S+/PATH/ > +--replace_regex /^\/\S+/PATH/ /^[a-zA-Z]:[\\\/]\S+/PATH/ /^BIGINT\b/<INT or BIGINT>/ /^INT\b/<INT or BIGINT>/ /^(?:18446744073709551615\b|4294967295\b)/<32-bit or 64-bit MAX UNSIGNED>/ /^(?:9223372036854775807\b|2147483647\b)/<32-bit or 64-bit MAX SIGNED>/ /^(?:18446744073709547520\b|4294963200\b)/<32-bit or 64-bit MAX UNSIGNED ADJUSTED>/ /^(?:9223372036853727232\b|2146435072\b)/<32-bit or 64-bit MAX SIGNED ADJUSTED>/ Aren't you removing too much with these replacements? It's kinda good to get rid of 32bit diff files (as they aren't easy to update), but I wouldn't replace every INT or BIGINT with <INT or BIGINT>. May be, move them to a separate select, like let type=BIGINT; if (32-bit) { let type=INT; } --replace $type <INT-or-BIGINT> select * from information_schema.system_variables where variable_name in (list of ulong variables); and something similar for other replacements. In fact, there's obscure feature of let $repl=/foo/bar/ /abc/def; replace_regex $repl; so you can have sysvar32bitrepl.inc that will set up $replace_regexes variable: let type=BIGINT; let max=18446744073709551615; if (32-bit) { let type=INT; let max=4294967295; } let replace_regexes=/$type/INT or BIGINT/ /$max/<32-bit or 64-bit MAX UNSIGNED>/; Then in every sysvar*.test file you can do, like --source sysvar32bitrepl --replace_regex $replace_regexes select * from information_schema.system_variables where variable_name in (list of ulong variables); Regards, Sergei

2 2

Re: [Maria-developers] MDEV-8389 patch review
by Sergey Petrunia 06 Aug '15

06 Aug '15

Hi Diwas, On Wed, Aug 05, 2015 at 08:47:28PM +0530, Diwas Joshi wrote: > I have removed the errors from previous patch, I only got those errors when > i compiled with valgrind. Everything seems to be working fine now. Also I > have fixed the test case for select * from mysql.proc. Also added a > function that removes the tables after they have been used. I have added > the errors that we discussed about and also test cases corresponding to > these errors. Please review. > The big problems in the patch are: 1. check_and_process_table_functions() is at the wrong place. I debug a query, "SELECT * FROM tbl" , where 'tbl' is a regular table. I see check_and_process_table_functions() invoke sp_find_routine() for "test.tbl". Why does this happen? The query makes it clear that the table is not a table function. This is not acceptable. 2. Temporary table 2.1 The temporary table that is created uses InnoDB storage engine. This doesn't follow any other kind of temporary table. When execution engine creates temporary tables (e.g. for VIEWs, or GROUP BY), these temporary tables are either Aria or MyISAM tables, depending on the server settings. We should follow that. 2.2 Temporary table is created by building the "CREATE TEMPORARY TABLE (...)" string and passing it to the parser. Did you try looking into creating the temp. table directly from the data strucutres you have? No other code has to build "CREATE TABLE" statements and then parse them back. I suspect Table Functions don't need to do this either. The rest of the feedback is below: > diff --git a/mysql-test/t/sp-error.test b/mysql-test/t/sp-error.test > index ccf5b98..2ad4144 100644 > --- a/mysql-test/t/sp-error.test > +++ b/mysql-test/t/sp-error.test > @@ -3846,7 +3846,23 @@ DROP FUNCTION f1; > DROP FUNCTION f2; > DROP TABLE t1; > > ---error 1235 > -SELECT * FROM f1('xyz'); > ---error 1064 > -SELECT * FROM t1, f1('xyz',t1.a); > +delimiter |; > +CREATE FUNCTION f1() RETURNS INTEGER > +BEGIN > + RETURN 1; > +END| > +delimiter ;| > +--error 1979 Please use error names, not error numbers. you can write --error ER_SP_SCALAR_VALUE_EXPECTED. > +SELECT * FROM f1(); > +DROP FUNCTION f1; > + > +delimiter |; > +CREATE FUNCTION f1(a INT, b VARCHAR(11)) > +RETURNS TABLE t1(id INT, name VARCHAR(11)) > +BEGIN > +INSERT INTO t1 SELECT id, name FROM t2 WHERE id = a; > +END| > +delimiter ;| > +--error 1978 Same as above. > +SELECT f1(1,'AA'); > +DROP FUNCTION f1; > diff --git a/mysql-test/t/sp.test b/mysql-test/t/sp.test > index 5c42df0..45db6a3 100644 > --- a/mysql-test/t/sp.test > +++ b/mysql-test/t/sp.test > @@ -9398,6 +9398,6 @@ END| > DELIMITER ;| > > --vertical_results > -SELECT * FROM mysql.proc WHERE db = database() and name = 'f1'; > - > +SELECT db, name, type, specific_name, language, sql_data_access, is_deterministic, security_type, param_list, returns, body, definer, sql_mode, comment, character_set_client, collation_connection, db_collation, body_utf8 FROM mysql.proc WHERE db = database() and name = 'f1'; The change is ok. Please also format the query so that its width is limited 80 chars. > +SELECT * FROM f1(10,'XYZ'); > DROP FUNCTION f1; > diff --git a/sql/item_func.cc b/sql/item_func.cc > index 3e78be8..ad18e2e 100644 > --- a/sql/item_func.cc > +++ b/sql/item_func.cc > @@ -6574,7 +6574,12 @@ void my_missing_function_error(const LEX_STRING &token, const char *func_name) > context->process_error(thd); > DBUG_RETURN(TRUE); > } > - > + > + if (m_sp->m_type == TYPE_ENUM_TABLE){ > + my_error(ER_SP_SCALAR_VALUE_EXPECTED, MYF(0)); > + context->process_error(thd); > + DBUG_RETURN(TRUE); > + } > /* > A Field need to be attached to a Table. > Below we "create" a dummy table by initializing > @@ -6588,7 +6593,7 @@ void my_missing_function_error(const LEX_STRING &token, const char *func_name) > dummy_table->copy_blobs= TRUE; > share->table_cache_key = empty_name; > share->table_name = empty_name; > - > + Please don't make dummy changes like this. > if (!(sp_result_field= m_sp->create_result_field(max_length, name, > dummy_table))) > { > diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc > index 2998972..6f94720 100644 > --- a/sql/sql_parse.cc > +++ b/sql/sql_parse.cc > @@ -2467,6 +2467,8 @@ static bool do_execute_sp(THD *thd, sp_head *sp) > /* should be assigned after making first tables same */ > all_tables= lex->query_tables; > /* set context for commands which do not use setup_tables */ > + if (lex->sql_command == SQLCOM_SELECT) > + check_and_process_table_functions(thd, all_tables); Why limit table function uses to SELECTs ? A user can write an INSERT.. SELECT statement INSERT INTO table SELECT * FROM table_func('argument'); or a multi-table UPDATE statement: UPDATE regular_table, table_func('arg') SET regular_table.col= ...; In general, there should not be a limit. I think, it's ok to introduce the limit within the scope of this MDEV entry. In that case, you need to add a comment with 'TODO-DiwasJoshi:...' saying that the limitation is temporary. > select_lex-> > context.resolve_in_table_list_only(select_lex-> > table_list.first); > @@ -5697,7 +5699,8 @@ static bool do_execute_sp(THD *thd, sp_head *sp) > thd->mdl_context.release_transactional_locks(); > } > #endif /* WITH_WSREP */ > - > + if (lex->sql_command == SQLCOM_SELECT) > + clear_table_functions(thd, all_tables); This looks like an odd place to do cleanups. Let's get back to this once other issues are resolved. > DBUG_RETURN(res || thd->is_error()); > } > > @@ -9222,3 +9225,185 @@ CHARSET_INFO *find_bin_collation(CHARSET_INFO *cs) > } > return cs; > } > + > +/* > + @brief > + Check for table functions and create corresponding tables to be used in query. > + > + @param thd Thread handle > + @param table_list A list of tables being processed in the query. > + > + @return > + False on success > + @retval > + TRUE on error > +*/ > +bool check_and_process_table_functions(THD *thd, TABLE_LIST *table_list) > +{ > + TABLE_LIST *tables = table_list; > + int err= 0; > + while(tables) > + { > + if (tables->table_name == NULL) > + { > + tables = tables->next_local; > + continue; > + } > + sp_head *sp; > + sp_name *name; > + LEX_STRING lex_db; > + LEX_STRING lex_name; > + lex_db.length= strlen(tables->db); > + lex_name.length= strlen(tables->table_name); > + lex_db.str= thd->strmake(tables->db, lex_db.length); > + lex_name.str= thd->strmake(tables->table_name, lex_name.length); > + name= new sp_name(lex_db, lex_name, true); > + name->init_qname(thd); > + int err=0; > + > + if (!err && (sp = sp_find_routine(thd,TYPE_ENUM_FUNCTION, name, > + &thd->sp_proc_cache, FALSE))) //check if the table is a table-valued function. > + { > + if (sp->m_type == TYPE_ENUM_FUNCTION) > + my_error(ER_SP_TABLE_VALUE_EXPECTED, MYF(0)); > + LEX *old_lex= thd->lex, newlex; > + thd->lex= &newlex; > + ulonglong sql_mode= 0 , saved_mode= thd->variables.sql_mode; > + String defstr; //query in defstr will be parsed to create temporary table. > + defstr.append(STRING_WITH_LEN("CREATE TEMPORARY TABLE ")); > + defstr.append(tables->table_name); > + defstr.append(STRING_WITH_LEN(" (")); > + List_iterator_fast<Create_field> it(sp->m_cols_list); > + > + TABLE table; > + TABLE_SHARE share; > + bzero((char*) &table, sizeof(table)); > + bzero((char*) &share, sizeof(share)); > + table.in_use= thd; > + table.s = &share; > + Create_field *cols_field; > + cols_field= it++; > + while (cols_field) > + { > + uint unused1= 0; > + sp_prepare_create_field(thd, cols_field); > + prepare_create_field(cols_field, &unused1, HA_CAN_GEOMETRY); > + uint field_length; > + Field *field; > + String tmp_result(512); > + field_length= cols_field->length; > + cols_field->flags= 0; > + field= make_field(table.s, /* TABLE_SHARE ptr */ > + (uchar*) 0, /* field ptr */ > + field_length, /* field [max] length */ > + (uchar*) "", /* null ptr */ > + 0, /* null bit */ > + cols_field->pack_flag, > + cols_field->sql_type, > + cols_field->charset, > + cols_field->geom_type, cols_field->srid, > + Field::NONE, /* unreg check */ > + cols_field->interval, > + cols_field->field_name); > + field->vcol_info= cols_field->vcol_info; > + field->stored_in_db= cols_field->stored_in_db; > + if (field) > + field->init(&table); > + defstr.append(cols_field->field_name); > + defstr.append(STRING_WITH_LEN(" ")); > + field->sql_type(tmp_result); > + defstr.append(tmp_result); > + cols_field= it++; > + if(cols_field) > + defstr.append(STRING_WITH_LEN(", ")); > + delete field; > + } > + defstr.append(STRING_WITH_LEN(")")); > + newlex.create_info.DDL_options_st::init(); > + newlex.current_select= NULL; > + Parser_state parser_state; > + thd->variables.sql_mode= sql_mode; > + err= parser_state.init(thd, defstr.c_ptr_safe(), defstr.length()); > + if (!err) { > + lex_start(thd); > + err= parse_sql(thd, &parser_state, NULL, TRUE) || thd->lex == NULL; > + } > + if (!err) > + { > + Table_specification_st create_info(thd->lex->create_info); > + Alter_info alter_info(thd->lex->alter_info, thd->mem_root); > + thd->lex->create_info.default_table_charset= NULL; > + SELECT_LEX *select_lex= &thd->lex->select_lex; > + TABLE_LIST *create_table= select_lex->table_list.first; > + if (append_file_to_dir(thd, &create_info.data_file_name, > + create_table->table_name) || > + append_file_to_dir(thd, &create_info.index_file_name, > + create_table->table_name)) > + { > + err= 1; > + } > + if (!(create_info.used_fields & HA_CREATE_USED_ENGINE)) > + create_info.use_default_db_type(thd); > + > + err= mysql_create_table(thd, create_table, &create_info, &alter_info); > + thd->variables.option_bits|= OPTION_KEEP_LOG; > + } > + thd->variables.sql_mode= saved_mode; > + lex_end(thd->lex); > + thd->lex= old_lex; > + } > + tables = tables->next_local; > + } > + if (!err) > + close_thread_tables(thd); > + return err; > +} > + > +/* > + @brief > + Drop tables created for table functions. > + > + @param thd Thread handle > + @param table_list A list of tables being processed in the query. > + > + @return > + False on success > + @retval > + TRUE on error > +*/ > +bool clear_table_functions(THD *thd, TABLE_LIST *table_list) > +{ > + TABLE_LIST *tables = table_list; > + int err= 0; > + while(tables) > + { > + if (tables->table_name == NULL) > + { > + tables = tables->next_local; > + continue; > + } > + sp_head *sp; > + sp_name *name; > + LEX_STRING lex_db; > + LEX_STRING lex_name; > + lex_db.length= strlen(tables->db); > + lex_name.length= strlen(tables->table_name); > + lex_db.str= thd->strmake(tables->db, lex_db.length); > + lex_name.str= thd->strmake(tables->table_name, lex_name.length); > + name= new sp_name(lex_db, lex_name, true); > + name->init_qname(thd); > + int err=0; > + if (!err && (sp = sp_find_routine(thd,TYPE_ENUM_FUNCTION, name, > + &thd->sp_proc_cache, FALSE)) && This function is supposed to just close the temporary tables. Why does it need to search for SPs? > + sp->m_type== TYPE_ENUM_TABLE) //check if the table is a table-valued function. > + { > + if(tables->table) > + close_temporary_table(thd,tables->table,TRUE,TRUE); > + if (thd->is_error()) > + err= 1; > + tables->table= NULL; > + } > + tables= tables->next_local; > + } > + return err; > +} BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] Best way of starting multi-source connections with active slave.
by Giuseppe Maxia 06 Aug '15

06 Aug '15

Hi, I am setting up multi-source replication with four nodes, where nodes 1, 2, and 3 are masters and node 4 is the slave of the first 3. The slave is *active*, meaning that it has a binary log and runs transactions of its own. All is well if I run 'reset master' in all nodes before setting up replication. CHANGE MASTER 'node1' TO master_host='host1', master_port=3306, master_user='some_slave', master_password='some_slave', master_use_gtid=current_pos; CHANGE MASTER 'node2' TO master_host='host2', master_port=3306, master_user='some_slave', master_password='some_slave', master_use_gtid=current_pos; CHANGE MASTER 'node3' TO master_host='host3', master_port=3306, master_user='some_slave', master_password='some_slave', master_use_gtid=current_pos; start all slaves; However, if host 4 has a binary log and has logged one or more transactions before starting replication, we get a conflict. Got fatal error 1236 from master when reading data from binary log: 'Error: connecting slave requested to start from GTID 0-104-1, which is not in the master's binlog. Since the master's binlog contains GTIDs with higher sequence numbers, it probably means that the slave has diverged due to executing extra erroneous transactions' I get the same result if I use 'slave_pos' instead of 'current_pos'. Notice that the changes in the slave do not conflict with anything that happens in the masters. For example: #1 node4 (fan-in slave): create user distribution_user identified by 'something'; #2 node4 CHANGE MASTER ... #3 node4 start all slaves #4 node1 (master): create table t1 #5 node2 (master): create table t2 #6 node3 (master): create table t3 This sequence of events fails. If I execute event #1 after #2 and #3, then all works as expected. Or it all works if I say master_use_gtid=no, but that would defeat the purpose of having GTIDs, or am I missing something? Is there a way of enabling multi-source replication with the slave having non-conflicting transactions already logged? Thanks Giuseppe

3 3

Re: [Maria-developers] 3eceb51: MDEV-7990: ERROR 1526 when procedure executed for second time ALTER TABLE partition ... pMAX values less than MAXVALUE
by Sergei Golubchik 05 Aug '15

05 Aug '15

Hi, Sanja! On May 06, sanja(a)mariadb.com wrote: > revision-id: 3eceb51e697d5a514e70c374a06e7eef6b5a37a0 > parent(s): aa5095627e2619bdad7916d33d1016802a84a9e1 > committer: Oleksandr Byelkin > branch nick: server > timestamp: 2015-05-06 13:19:22 +0200 > message: > > MDEV-7990: ERROR 1526 when procedure executed for second time ALTER > TABLE partition ... pMAX values less than MAXVALUE > > Added cleaning partition data which are saved in LEX structure to make > possible correct re-execution of a SP or PS using the code. No, this is wrong. See prep_alter_part_table(). The first thing it does it to copy thd->lex->part_info to thd->work_part_info. It tries to do a deep copy to be able to modify it freely, but apparently the copy isn't deep enough. I see two options 1. Do a deeper copy. Duplicate all value lists. 2. Make part_column_list_val elements immutable. That would be preferred, of course, but I'm not sure it's possible. Regards, Sergei

2 2