- developers - lists.mariadb.org

Re: 2342aaf46eb: MDEV-30973 Implement debug logging for slave_transaction_retries
by Sergei Golubchik 07 Nov '23

07 Nov '23

Hi, Aleksey, On Nov 07, Aleksey Midenkov wrote: > revision-id: 2342aaf46eb (mariadb-11.0.1-256-g2342aaf46eb) > parent(s): 80bc5df1c80 > author: Aleksey Midenkov > committer: Aleksey Midenkov > timestamp: 2023-10-10 13:11:19 +0300 > message: > > MDEV-30973 Implement debug logging for slave_transaction_retries > > Configuration options: > > log_slave_retries > log_slave_retries_max 1. a commit comment of >100 lines that's a bit too much, I think. 2. comments in MDEV-30973 say that one should be able to have everything in one file, namely in the error log. I don't see how it's possible in your commit. 3. one of the reasons for not simply writing into erorr log was "you get what are relatively large entries in there which can make it hard/annoying to also look for the normal, one-line warning and error output". But as far as I can see your slave retry entries are not large, but normal one-line messages. I'd say they can fit into the error log just fine. I've still commented below in the assumption that the separate log implementation will stay. But I honestly don't see a reason for it anymore. > diff --git a/mysql-test/suite/rpl/t/slave_deadlock.test b/mysql-test/suite/rpl/t/slave_deadlock.test > new file mode 100644 > index 00000000000..9fcda7c9b71 > --- /dev/null > +++ b/mysql-test/suite/rpl/t/slave_deadlock.test > @@ -0,0 +1,194 @@ > +--source include/have_debug.inc > +--source include/have_innodb.inc > +--source include/have_binlog_format_statement.inc > + > +# On --repeat test infinitely increments thread_id and finally fails --sync_with_master at log_pos 1754: > +# > +# 231009 20:58:15 server id 1 end_log_pos 1754 CRC32 0x7e5aecd1 Query thread_id=561 exec_time=0 error_code=0 xid=0 I'm sorry, I don't understand that. Could you elaborate please? > +# > +# force_restart seems to fix this > + > +--source include/force_restart.inc > + > +# Note that test requires big enough --debug-sync-timeout unfair scheduling at high load. Sorry, cannot parse that. Could you elaborate please? > +# slave_deadlock.opt sets this to 1 hour > + > +--let $rpl_topology=1->2 > +--source include/rpl_init.inc > + > +--echo # > +--echo # MDEV-30973 Enhance output of innodb deadlocks (SAMU-124, SAMU-131) > +--echo # > +--connection server_2 > +SET @old_dbug= @@global.debug_dbug; > +set @old_parallel_threads= @@global.slave_parallel_threads; > +set @old_domain_parallel_threads= @@global.slave_domain_parallel_threads; > +set @old_parallel_mode= @@global.slave_parallel_mode; > +set @old_slave_transaction_retries= @@global.slave_transaction_retries; > +set @old_innodb_lock_wait_timeout= @@global.innodb_lock_wait_timeout; > +set @old_log_slave_retries= @@global.log_slave_retries; > +set @old_log_slave_retries_file= @@global.log_slave_retries_file; > +set @old_log_slave_retries_max= @@global.log_slave_retries_max; All this makes little sense, it's simpler to put all settings into the .opt file generally it's better to do your way, to have less server restarts, but you have an .opt file already, you can as well use it. except if you decide to remove the .opt file - then nevermind, the above it good. > +--source include/stop_slave.inc > +change master to master_use_gtid= slave_pos; > +set global slave_parallel_threads= 10; > +set global slave_domain_parallel_threads= 1; > +set global slave_parallel_mode= optimistic; > +set global slave_transaction_retries= 4; > +set global innodb_lock_wait_timeout= 1; ... > diff --git a/mysql-test/suite/versioning/r/rpl_row.result b/mysql-test/suite/versioning/r/rpl_row.result > index c161f68d0bd..6343fbecdec 100644 > --- a/mysql-test/suite/versioning/r/rpl_row.result > +++ b/mysql-test/suite/versioning/r/rpl_row.result > @@ -37,12 +37,7 @@ select *, unix_timestamp(row_start) as row_start, unix_timestamp(row_end) as row > x y row_start row_end > 1 1 12345.000000 2147483647.999999 > set timestamp= default; > -### INSERT INTO `test`.`t1` > -### SET > -### @1=1 > -### @2=1 > -### @3=12345.000000 > -### @4=2147483647.999999 why? > +/*!50530 SET @@SESSION.PSEUDO_SLAVE_MODE=1*/; > connection slave; > select * from t1; > x y > diff --git a/sql/log.cc b/sql/log.cc > index 877315c60de..78e874fda5b 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -1188,9 +1188,17 @@ bool Log_to_file_event_handler:: > log_error(enum loglevel level, const char *format, > va_list args) > { > - return vprint_msg_to_log(level, format, args); > + return vprint_msg_to_log(stderr, level, format, args); > } > > +#ifdef HAVE_REPLICATION > +bool Log_to_file_event_handler::log_slave_retry(const char *format, va_list args) > +{ > + DBUG_ASSERT(slave_retries_file.get()); > + return vprint_msg_to_log(slave_retries_file.get(), NO_LEVEL, format, args); I don't think you should use vprint_msg_to_log and add a NO_LEVEL hack. vprint_msg_to_log also prints into the windows event log, into wsrep json log, and it locks LOCK_error_log. You don't need any of that, just create your own function, it's just one fprintf() anyway. > +} > +#endif > + > void Log_to_file_event_handler::init_pthread_objects() > { > mysql_log.init_pthread_objects(); > @@ -9752,13 +9787,15 @@ static void print_buffer_to_file(enum loglevel level, const char *buffer, > } > } > > +#ifndef EMBEDDED_LIBRARY > mysql_mutex_lock(&LOCK_error_log); > +#endif > > skr= my_time(0); > localtime_r(&skr, &tm_tmp); > start=&tm_tmp; > > - fprintf(stderr, "%d-%02d-%02d %2d:%02d:%02d %lu [%s] %.*s%.*s\n", > + fprintf(file, "%d-%02d-%02d %2d:%02d:%02d [T%lu] %s%.*s%.*s\n", eh... no. don't change the format of existing log files, please. > start->tm_year + 1900, > start->tm_mon+1, > start->tm_mday, > @@ -9834,6 +9872,22 @@ void sql_print_error(const char *format, ...) > } > > > +#ifdef HAVE_REPLICATION > +void sql_print_error2(const char *format, ...) s/sql_print_error2/print_slave_error/ ? > +{ > + va_list args; > + DBUG_ENTER("sql_print_error"); > + > + va_start(args, format); > + error_log_print(ERROR_LEVEL, format, args); > + slave_retries_print(format, args); > + va_end(args); > + > + DBUG_VOID_RETURN; > +} > +#endif > + > + > void sql_print_warning(const char *format, ...) > { > va_list args; > diff --git a/sql/mysqld.cc b/sql/mysqld.cc > index 8cc3ad51ee9..4f6b2273597 100644 > --- a/sql/mysqld.cc > +++ b/sql/mysqld.cc > @@ -765,6 +765,90 @@ char *opt_relay_logname = 0, *opt_relaylog_index_name=0; > char *opt_logname, *opt_slow_logname, *opt_bin_logname; > char *opt_binlog_index_name=0; > > +#ifdef HAVE_REPLICATION > +char default_slave_retries_path[FN_REFLEN]; > +bool opt_slave_retries= false; > +char *opt_slave_retries_path= NULL; > +uint opt_slave_retries_max_log= 0; > +#endif > +Slave_retries_file slave_retries_file; > + > +bool Slave_retries_file::open() > +{ > +#ifdef HAVE_REPLICATION > + DBUG_ASSERT(opt_slave_retries); > + DBUG_ASSERT(opt_slave_retries_path); > + file= fopen(opt_slave_retries_path, "a"); my_fopen or my_open and anyway, why a new class, why not to use MYSQL_LOG like general and slow logs do? > + if (!file) > + { > + sql_print_error("Open of log_slave_retries '%s' failed: %s (%d)", > + opt_slave_retries_path, strerror(errno), errno); This is inconsistent, the server doesn't print this for other log files. I've just tried with general log and got 2023-11-06 17:58:19 0 [ERROR] mysqld: File '/tmp' not found (Errcode: 21 "Is a directory") 2023-11-06 17:58:19 0 [ERROR] Could not use /tmp for logging (error 21). Turning logging off for the whole duration of the MariaDB server process. To turn it on again: fix the cause, shutdown the MariaDB server and restart it. Please, do the same > + return true; > + } > + else > + { > + slave_retries_print("%s (mariadbd %s) starting as process %lu ...", > + my_progname, server_version, (ulong) getpid()); Any reason, why it shouldn't be the same as in the error log? sql_print_information("Starting MariaDB %s source revision %s as process %lu", server_version, SOURCE_REVISION, (ulong) getpid()); > + } > +#endif > + return false; > +} > + > +void Slave_retries_file::close() > +{ > +#ifdef HAVE_REPLICATION > + if (file) > + { > + acquire(); > + fclose(file); > + release(); > + } > + file= NULL; > +#endif > +} > + > +void Slave_retries_file::init() > +{ > +#ifdef HAVE_REPLICATION > + mysql_mutex_init(key_LOCK_global_system_variables, &mutex, MY_MUTEX_INIT_FAST); > +#endif > +} > + > +void Slave_retries_file::destroy() > +{ > +#ifdef HAVE_REPLICATION > + mysql_mutex_destroy(&mutex); > +#endif > +} > + > +bool Slave_retries_file::acquire() > +{ > +#ifdef HAVE_REPLICATION > + if (!file) > + return false; > + mysql_mutex_lock(&mutex); > + return true; > +#else > + return false; > +#endif > +} > + > +void Slave_retries_file::release() > +{ > +#ifdef HAVE_REPLICATION > + mysql_mutex_unlock(&mutex); > +#endif > +} > + > +void Slave_retries_file::flush() > +{ > + if (acquire()) > + { > + fflush(file); > + release(); > + } > +} > + > /* Static variables */ > > my_bool opt_stack_trace; > @@ -5055,6 +5145,35 @@ static int init_server_components() > "this server. However this will be ignored as the " > "--log-bin option is not defined."); > } > + > + slave_retries_file.init(); > + if (!log_error_file_ptr[0]) > + fn_format(default_slave_retries_path, pidfile_name, mysql_data_home, > + "-retries.err", MY_REPLACE_EXT); > + else > + fn_format(default_slave_retries_path, log_error_file_ptr, mysql_data_home, > + "-retries.err", MY_REPLACE_EXT); that's not how all other log names work, grep mysqld.cc for opt_slow_logname > + if (opt_slave_retries_path && !opt_abort) > + { > + if (!opt_slave_retries_path[0]) > + { > + if (!default_slave_retries_path[0]) > + { > + /* fn_format failed */ > + opt_slave_retries_path= NULL; > + opt_slave_retries= false; > + } > + else > + { > + opt_slave_retries_path= my_strdup(key_memory_Sys_var_charptr_value, > + default_slave_retries_path, MYF(0)); > + } > + } > + > + if (opt_slave_retries && > + slave_retries_file.open()) > + unireg_abort(1); > + } > #endif > > if (opt_bin_log) > diff --git a/sql/mysqld.h b/sql/mysqld.h > index 69cfc0cad56..5f1561ca10d 100644 > --- a/sql/mysqld.h > +++ b/sql/mysqld.h > @@ -822,6 +848,8 @@ enum options_mysqld > OPT_KEY_CACHE_CHANGED_BLOCKS_HASH_SIZE, > OPT_LOG_BASENAME, > OPT_LOG_ERROR, > + OPT_SLAVE_RETRIES, > + OPT_SLAVE_RETRIES_LOG, why do you need them? > OPT_LOWER_CASE_TABLE_NAMES, > OPT_PLUGIN_LOAD, > OPT_PLUGIN_LOAD_ADD, > diff --git a/sql/sql_show.cc b/sql/sql_show.cc > index d742bc6c1c3..c53ca0f0097 100644 > --- a/sql/sql_show.cc > +++ b/sql/sql_show.cc > @@ -10662,6 +10663,15 @@ char *thd_get_error_context_description(THD *thd, char *buffer, > /* Don't wait if LOCK_thd_data is used as this could cause a deadlock */ > if (!mysql_mutex_trylock(&thd->LOCK_thd_data)) > { > + if (thd->rgi_slave) is rgi_slave and its content protected by thd->LOCK_thd_data ? > + { > + rpl_group_info *rgi= thd->rgi_slave; > + len= my_snprintf(header, sizeof(header), > + ", GTID %u-%u-%llu", rgi->current_gtid.domain_id, > + rgi->current_gtid.server_id, rgi->current_gtid.seq_no); > + str.append(header, len); > + } > + > if (const char *info= thread_state_info(thd)) > { > str.append(' '); > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index ebb64b340b0..11c321f06dc 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -5862,6 +5862,61 @@ Sys_slave_trans_retry_interval( > "slave_transaction_retry_errors", > GLOBAL_VAR(slave_trans_retry_interval), CMD_LINE(REQUIRED_ARG), > VALID_RANGE(0, 3600), DEFAULT(0), BLOCK_SIZE(1)); > + > +static bool update_slave_retries_log(sys_var *self, THD* thd, enum_var_type type) > +{ > + if (opt_slave_retries) > + { > + if (!slave_retries_file.get() && slave_retries_file.open()) > + { > + my_error(ER_CANT_OPEN_FILE, MYF(0), opt_slave_retries_path, errno); > + opt_slave_retries= false; > + return true; > + } > + } > + else > + slave_retries_file.close(); > + return false; > +} > + > +static Sys_var_mybool Sys_slave_retries_log( > + "log_slave_retries", "Log transaction retries", > + GLOBAL_VAR(opt_slave_retries), > + CMD_LINE(OPT_ARG, OPT_SLAVE_RETRIES), DEFAULT(false), > + NO_MUTEX_GUARD, NOT_IN_BINLOG, ON_CHECK(0), ON_UPDATE(update_slave_retries_log)); > + > + > +static bool check_slave_retries_file(sys_var *self, THD *thd, set_var *var) > +{ > + if (opt_slave_retries) > + { > + my_error(ER_NOT_SUPPORTED_YET, MYF(0), "modifying while log_slave_retries is ON"); as always, please don't hard-code error messages. you can use my_error(ER_OPTION_PREVENTS_STATEMENT, MYF(0), "log_slave_retries=ON"); > + return true; > + } > + return false; > +} > + > +static bool update_slave_retries_file(sys_var *self, THD* thd, enum_var_type type) > +{ > + if (!opt_slave_retries_path) > + opt_slave_retries_path= my_strdup(key_memory_Sys_var_charptr_value, > + default_slave_retries_path, MYF(0)); sys_var.cc can handle that for you, see PREALLOCATED in e.g. opt_slow_logname also, you don't reopen the log when its name changes, how comes? > + return false; > +} > + > +static Sys_var_charptr Sys_slave_retries_log_path( > + "log_slave_retries_file", "Log file path for transaction retries", > + GLOBAL_VAR(opt_slave_retries_path), > + CMD_LINE(REQUIRED_ARG, OPT_SLAVE_RETRIES_LOG), DEFAULT(NULL), > + NO_MUTEX_GUARD, NOT_IN_BINLOG, ON_CHECK(check_slave_retries_file), > + ON_UPDATE(update_slave_retries_file)); > + > + > +static Sys_var_uint Sys_slave_retries_log_max( > + "log_slave_retries_max", "Log max N transaction retries " > + "(default 0 means infinity)", just "(0 means infinity)". Default is printed separately > + GLOBAL_VAR(opt_slave_retries_max_log), CMD_LINE(REQUIRED_ARG), > + VALID_RANGE(0, UINT_MAX), DEFAULT(0), BLOCK_SIZE(1)); > #endif > > static bool check_locale(sys_var *self, THD *thd, set_var *var) > diff --git a/storage/innobase/lock/lock0lock.cc b/storage/innobase/lock/lock0lock.cc > index 0e9cef669d4..8a8eeaf39f6 100644 > --- a/storage/innobase/lock/lock0lock.cc > +++ b/storage/innobase/lock/lock0lock.cc > @@ -49,6 +49,7 @@ Created 5/7/1996 Heikki Tuuri > #include <debug_sync.h> > > #include <set> > +#include <stdio.h> I presume Marko has reviewed (or will review) InnoDB changes > > #ifdef WITH_WSREP > #include <mysql/service_wsrep.h> > diff --git a/mysql-test/main/mysqld--help.result b/mysql-test/main/mysqld--help.result > index 77d6d10b3e6..274c034d455 100644 > --- a/mysql-test/main/mysqld--help.result > +++ b/mysql-test/main/mysqld--help.result > @@ -1673,6 +1678,9 @@ log-isam myisam.log > log-output FILE > log-queries-not-using-indexes FALSE > log-short-format FALSE > +log-slave-retries FALSE > +log-slave-retries-file (No default value) this is another sign that log-slave-retries-file doesn't behave like other log filename variables. > +log-slave-retries-max 0 > log-slave-updates FALSE > log-slow-admin-statements TRUE > log-slow-disabled-statements sp Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: [PATCH] MDEV-32673 Handle InnoDB binlog position recovery for RESET MASTER or binlog name change
by Kristian Nielsen 07 Nov '23

07 Nov '23

Hi Marko, Thanks for the patch. It looks good, just a few small comments inline. Approved after fixing the check for !is_relay_log. I have pushed an extra commit on top of bb-10.4-MDEV-32673 with a testcase (feel free to squash it onto your commit). The tests pass with your patch (but fails without it). - Kristian. > diff --git a/sql/handler.h b/sql/handler.h > index 6085111bc25..134212a0ad7 100644 > --- a/sql/handler.h > +++ b/sql/handler.h > @@ -1438,6 +1438,7 @@ struct handlerton > enum_binlog_command binlog_command, > const char *query, uint query_length, > const char *db, const char *table_name); > + void (*reset_binlog_pos)(const char *name, uint64_t pos); I think there should be a comment describing what this new call does. Suggestion: /* Called at server startup or RESET MASTER when the new binlog is opened. Informs the storage engine of the change in binlog file name and position. */ > diff --git a/sql/log.cc b/sql/log.cc > index 59730c3205a..e46f3796b63 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -4327,6 +4327,8 @@ bool MYSQL_BIN_LOG::reset_logs(THD *thd, bool create_new_log, > err: > if (error == 1) > name= const_cast<char*>(save_name); > + else > + ha_reset_binlog_pos(get_log_fname(), BIN_LOG_HEADER_SIZE); Sorry, I forgot something when we discussed this. This function is also used for resetting the relay logs. So it needs a check for !is_relay_log. For example: --- a/sql/log.cc +++ b/sql/log.cc @@ -4331,6 +4331,10 @@ bool MYSQL_BIN_LOG::reset_logs(THD *thd, bool create_new_log, if (!is_relay_log) { xid_count_per_binlog *b; + + if (!error) + ha_reset_binlog_pos(get_log_fname(), BIN_LOG_HEADER_SIZE); + /* Remove all entries in the xid_count list except the last. Normally we will just be deleting all the entries that we waited for to > diff --git a/storage/innobase/trx/trx0rseg.cc b/storage/innobase/trx/trx0rseg.cc > index be52373d03b..3ffa003588d 100644 > --- a/storage/innobase/trx/trx0rseg.cc > +++ b/storage/innobase/trx/trx0rseg.cc > @@ -826,3 +826,52 @@ trx_rseg_update_binlog_offset(byte* rseg_header, const trx_t* trx, mtr_t* mtr) > mlog_write_string(p, binlog_name, len, mtr); > } > } > + > +void trx_rseg_reset_binlog_pos(const char *name, uint64_t pos) > +{ > + mtr_t mtr; > + /* Invalidate the information on the TRX_SYS page. */ > + mtr.start(); > + if (buf_block_t *sys= trx_sysf_get(&mtr, true)) > + { > + byte *magic= TRX_SYS + TRX_SYS_MYSQL_LOG_INFO + > + TRX_SYS_MYSQL_LOG_MAGIC_N_FLD + sys->frame; > + if (mach_read_from_4(magic)) > + mlog_memset(magic, > + TRX_SYS_MYSQL_LOG_NAME + TRX_SYS_MYSQL_LOG_NAME_LEN, 0, > + &mtr); > + > + const uint32_t page_no= trx_sysf_rseg_get_page_no(sys, 0); > + if (page_no != FIL_NULL) > + { > + const uint32_t space_id= trx_sysf_rseg_get_space(sys, 0); > + ut_ad(space_id == TRX_SYS_SPACE); > + buf_block_t *block= buf_page_get(page_id_t(space_id, page_no), > + 0, RW_X_LATCH, &mtr); > + mlog_write_ull(TRX_RSEG + TRX_RSEG_BINLOG_OFFSET + block->frame, > + pos, &mtr); > + byte *p= TRX_RSEG + TRX_RSEG_BINLOG_NAME + block->frame; > + const size_t len= strlen(name) + 1; > + if (memcmp(name, p, len)) > + mlog_write_string(p, reinterpret_cast<const byte*>(name), len, &mtr); > + } > + > + /* Invalidate the information in the other rollback segment headers. */ > + for (ulint rseg_id= 1; rseg_id < TRX_SYS_N_RSEGS; rseg_id++) > + { > + const uint32_t page_no= trx_sysf_rseg_get_page_no(sys, rseg_id); > + if (page_no != FIL_NULL) > + { > + const uint32_t space_id= trx_sysf_rseg_get_space(sys, rseg_id); > + buf_block_t *block= buf_page_get(page_id_t(space_id, page_no), > + 0, RW_X_LATCH, &mtr); > + byte *name= TRX_RSEG + TRX_RSEG_BINLOG_NAME + block->frame; > + if (*name) > + mlog_write_ulint(name, 0, MLOG_1BYTE, &mtr); > + } > + } > + } > + mtr.commit(); > + /* Durably write the change. */ > + log_write_up_to(mtr.commit_lsn(), true); > +}

1 0

Re: 3aea3bd7422: MDEV-23294 Segfault or assertion upon MyISAM repair
by Sergei Golubchik 06 Nov '23

06 Nov '23

Hi, Aleksey, On Oct 23, Aleksey Midenkov wrote: > revision-id: 3aea3bd7422 (mariadb-10.4.30-176-g3aea3bd7422) > parent(s): 4fa6383c937 > author: Aleksey Midenkov > committer: Aleksey Midenkov > timestamp: 2023-09-26 20:20:46 +0300 > message: > > MDEV-23294 Segfault or assertion upon MyISAM repair > > When computing vcol expression some items use current_thd and that was > not set in MyISAM repair thread. Since all the repair threads belong > to one connection and items should not write into THD we can utilize > table THD for that. > > diff --git a/storage/myisam/ha_myisam.cc b/storage/myisam/ha_myisam.cc > index ddab8bcf741..27151548704 100644 > --- a/storage/myisam/ha_myisam.cc > +++ b/storage/myisam/ha_myisam.cc > @@ -714,6 +714,8 @@ static int compute_vcols(MI_INFO *info, uchar *record, int keynum) > /* This mutex is needed for parallel repair */ > mysql_mutex_lock(&info->s->intern_lock); > TABLE *table= (TABLE*)(info->external_ref); > + if (!current_thd) > + set_current_thd(table->in_use); why not to do it in thr_find_all_keys(), unconditionally? you'd avoid an if() executed per row. Also, I'd suggest a comment, like /* To evaluate vcols we must have current_thd set. This will set current_thd in all threads to the same THD, but it's safe, because vcols are always evaluated under info->s->intern_lock. */ set_current_thd((TABLE*)(info->external_ref)->in_use); > table->move_fields(table->field, record, table->field[0]->record_ptr()); > if (keynum == -1) // update all vcols > { Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 3

Re: 0f0ef93c4d4: MDEV-32444 Data from orphaned XA transaction is lost after online alter
by Sergei Golubchik 02 Nov '23

02 Nov '23

Hi, Nikita, See below. A small suggestion for the test. And a rather larger comment about extending xid_t, as you, likely, expected :) On Oct 19, Nikita Malyavin wrote: > revision-id: 0f0ef93c4d4 (mariadb-11.2.1-15-g0f0ef93c4d4) > parent(s): 7f00853f2f1 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-10-13 13:39:36 +0400 > diff --git a/mysql-test/main/alter_table_online_debug.result b/mysql-test/main/alter_table_online_debug.result > index 1ebb19965eb..154aec003ca 100644 > --- a/mysql-test/main/alter_table_online_debug.result > +++ b/mysql-test/main/alter_table_online_debug.result > @@ -1563,6 +1563,103 @@ connection default; > drop table t1, t2; > set @@binlog_format=default; > set debug_sync= reset; > +# MDEV-32444 Data from orphaned XA transaction is lost after online alter > +create table t (a int primary key) engine=innodb; > +insert into t values (1); > +# XA commit > +set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go'; > +alter table t force, algorithm=copy, lock=none; > +connection con1; > +set debug_sync= 'now wait_for downgraded'; > +xa begin 'x1'; > +update t set a = 2 where a = 1; > +xa end 'x1'; > +xa prepare 'x1'; > +disconnect con1; > +connection con2; > +xa commit 'x1'; > +set debug_sync= 'now signal go'; > +connection default; > +select * from t; > +a > +2 > +# XA rollback > +set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go'; > +alter table t force, algorithm=copy, lock=none; > +connect con1, localhost, root,,; > +xa begin 'x2'; > +insert into t values (53); > +xa end 'x2'; > +xa prepare 'x2'; > +disconnect con1; > +connection con2; > +xa rollback 'x2'; > +set debug_sync= 'now signal go'; > +connection default; > +select * from t; > +a > +2 > +# XA transaction is left uncommitted > +# end then is rollbacked after alter fails > +set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go'; > +set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0 > +for alter table t force, algorithm=copy, lock=none; > +connect con1, localhost, root,,; > +xa begin 'xuncommitted'; > +insert into t values (3); > +xa end 'xuncommitted'; > +xa prepare 'xuncommitted'; > +set debug_sync= 'now signal go'; > +disconnect con1; > +connection default; > +ERROR HY000: Lock wait timeout exceeded; try restarting transaction > +xa rollback 'xuncommitted'; > +select * from t; > +a > +2 > +# Same, but commit > +set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go'; > +set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0 > +for alter table t force, algorithm=copy, lock=none; instead of `alter table t force`, better do something that changes the table, so that you could see whether alter was executed or aborted. For example, you can change the column type int->bigint and back. not only here, but everywhere. Or add/drop column, this has an added benefit that you won't need a separate `show create table`, your already existing `select * from t` will show whether the extra column exists > +connect con1, localhost, root,,; > +xa begin 'committed_later'; > +insert into t values (3); > +xa end 'committed_later'; > +xa prepare 'committed_later'; > +set debug_sync= 'now signal go'; > +disconnect con1; > +connection default; > +ERROR HY000: Lock wait timeout exceeded; try restarting transaction > +xa commit 'committed_later'; > +select * from t; > +a > +2 > +3 > +# Commit, but error in statement, and there is some stmt data to rollback > +set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go'; > +alter table t force, algorithm=copy, lock=none; > +connect con1, localhost, root,,; > +set debug_sync= 'now wait_for downgraded'; > +xa begin 'x1'; > +insert into t values (4), (3); > +ERROR 23000: Duplicate entry '3' for key 'PRIMARY' > +insert into t values (5); > +xa end 'x1'; > +xa prepare 'x1'; > +disconnect con1; > +connection con2; > +xa commit 'x1'; > +set debug_sync= 'now signal go'; > +connection default; > +select * from t; > +a > +2 > +3 > +5 > +connect con1, localhost, root,,; > +connection default; > +drop table t; > +set debug_sync= reset; > disconnect con1; > disconnect con2; > # > diff --git a/sql/handler.h b/sql/handler.h > index 07b392c010a..4622e49f883 100644 > --- a/sql/handler.h > +++ b/sql/handler.h > @@ -903,6 +905,7 @@ struct xid_t { > long gtrid_length; > long bqual_length; > char data[XIDDATASIZE]; // not \0-terminated ! > + Online_alter_cache_list *online_alter_cache; well, no, we've talked about it. Even the comment above the structure says that you cannot change it. make a separate structure, xid_and_online_alter_cache_t or xid_t_internal or whatever. but xid_t should be defined identically everywhere. For example: https://github.com/berkeleydb/libdb/blob/master/src/dbinc/xa.h === okay, now I've seen how you used it and I'd say it's _such a rare use case_ that you can store your online_alter_cache in the XID_cache_element and look it up there. Or even in a separate data structure, that will be empty most of the time anyway. > > xid_t() = default; /* Remove gcc warning */ > bool eq(struct xid_t *xid) const > diff --git a/sql/online_alter.cc b/sql/online_alter.cc > index 8b8c81319d1..19abe57914d 100644 > --- a/sql/online_alter.cc > +++ b/sql/online_alter.cc > @@ -334,14 +336,57 @@ static int online_alter_log_init(void *p) > online_alter_hton->savepoint_rollback_can_release_mdl= > [](handlerton *hton, THD *thd){ return true; }; > > - online_alter_hton->commit= [](handlerton *hton, THD *thd, bool all) > - { return online_alter_end_trans(hton, thd, all, true); }; > - online_alter_hton->rollback= [](handlerton *hton, THD *thd, bool all) > - { return online_alter_end_trans(hton, thd, all, false); }; > - online_alter_hton->commit_by_xid= [](handlerton *hton, XID *xid) > - { return online_alter_end_trans(hton, current_thd, true, true); }; > - online_alter_hton->rollback_by_xid= [](handlerton *hton, XID *xid) > - { return online_alter_end_trans(hton, current_thd, true, false); }; > + online_alter_hton->commit= [](handlerton *hton, THD *thd, bool all) -> int > + { > + int res= online_alter_end_trans(get_cache_list(hton, thd), thd, > + ending_trans(thd, all), true); > + cleanup_tables(thd); > + return res; > + }; > + online_alter_hton->rollback= [](handlerton *hton, THD *thd, bool all) -> int > + { > + int res= online_alter_end_trans(get_cache_list(hton, thd), thd, > + ending_trans(thd, all), false); > + cleanup_tables(thd); > + return res; > + }; > + > + > + online_alter_hton->recover= [](handlerton*, XID*, uint){ return 0; }; > + online_alter_hton->prepare= [](handlerton *hton, THD *thd, bool all) -> int > + { > + auto &cache_list= get_cache_list(hton, thd); > + int res= 0; > + if (ending_trans(thd, all)) > + { > + thd->transaction->xid_state.set_online_alter_cache(&cache_list); > + thd_set_ha_data(thd, hton, NULL); > + } > + else > + { > + res= online_alter_end_trans(cache_list, thd, false, true); is that possible? > + } > + > + cleanup_tables(thd); > + return res; > + }; > + online_alter_hton->commit_by_xid= [](handlerton *hton, XID *xid) -> int > + { > + int res= online_alter_end_trans(*xid->online_alter_cache, current_thd, > + true, true); > + delete xid->online_alter_cache; > + xid->online_alter_cache= NULL; > + return res; > + }; > + online_alter_hton->rollback_by_xid= [](handlerton *hton, XID *xid) -> int > + { > + int res= online_alter_end_trans(*xid->online_alter_cache, current_thd, > + true, false); > + delete xid->online_alter_cache; > + xid->online_alter_cache= NULL; > + return res; > + }; > + > > online_alter_hton->drop_table= [](handlerton *, const char*) { return -1; }; > online_alter_hton->flags= HTON_NOT_USER_SELECTABLE | HTON_HIDDEN Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 6

Re: 2d937b62c33: MDEV-27744 InnoDB: Failing assertion: !cursor->index->is_committed() in row0ins.cc (from row_ins_sec_index_entry_by_modify) | Assertion `0' failed in row_upd_sec_index_entry (debug) | Corruption
by Sergei Golubchik 02 Nov '23

02 Nov '23

Hi, Alexander, Few questions and suggestions below. Nothing major. On Oct 27, Alexander Barkov wrote: > commit 2d937b62c33 > Author: Alexander Barkov <bar(a)mariadb.com> > Date: Mon Apr 4 14:50:21 2022 +0400 > > MDEV-27744 InnoDB: Failing assertion: !cursor->index->is_committed() in row0ins.cc (from row_ins_sec_index_entry_by_modify) | Assertion `0' failed in row_upd_sec_index_entry (debug) | Corruption > would it make sense to change the MDEV title in Jira to better describe the problem? I sometimes do it with my bugs. Like "LPAD in vcol created in ORACLE mode makes table corrupted in non-ORACLE" (I tried to make it short, appropriate for a title) and, of course, change it in the comment and in the test file to match. > The crash happened with an indexed virtual column whose > value is evaluated using a function that has a different meaning > in sql_mode='' vs sql_mode=ORACLE: > > diff --git a/mysql-test/suite/compat/oracle/r/func_concat.result b/mysql-test/suite/compat/oracle/r/func_concat.result > index 392d579707a..17ca4be078a 100644 > --- a/mysql-test/suite/compat/oracle/r/func_concat.result > +++ b/mysql-test/suite/compat/oracle/r/func_concat.result > @@ -211,14 +211,14 @@ SET sql_mode=ORACLE; > CREATE VIEW v1 AS SELECT 'foo'||NULL||'bar' AS test; > SHOW CREATE VIEW v1; > View Create View character_set_client collation_connection > -v1 CREATE VIEW "v1" AS select concat_operator_oracle(concat_operator_oracle('foo',NULL),'bar') AS "test" latin1 latin1_swedish_ci > +v1 CREATE VIEW "v1" AS select concat(concat('foo',NULL),'bar') AS "test" latin1 latin1_swedish_ci > SELECT * FROM v1; > test > foobar > SET sql_mode=DEFAULT; > SHOW CREATE VIEW v1; > View Create View character_set_client collation_connection > -v1 CREATE ALGORITHM=UNDEFINED DEFINER=`root`@`localhost` SQL SECURITY DEFINER VIEW `v1` AS select concat_operator_oracle(concat_operator_oracle('foo',NULL),'bar') AS `test` latin1 latin1_swedish_ci > +v1 CREATE ALGORITHM=UNDEFINED DEFINER=`root`@`localhost` SQL SECURITY DEFINER VIEW `v1` AS select oracle_schema.concat(oracle_schema.concat('foo',NULL),'bar') AS `test` latin1 latin1_swedish_ci please, add a mysqldump test. Like, create a table with virtual columns, check, defaults in the default sql mode. and create a table in oracle mode. Also, a view in the default mode and a view in oracle mode. and then mysqldump, to see that it dumps and restores everything correctly may be stored routines/triggers/etc, if you'd like, but they aren't directly relevant to this MDEV, as far as I understand. > SELECT * FROM v1; > test > foobar > diff --git a/mysql-test/suite/compat/oracle/r/func_decode.result b/mysql-test/suite/compat/oracle/r/func_decode.result > index 2809e971be3..1870a1ec0d5 100644 > --- a/mysql-test/suite/compat/oracle/r/func_decode.result > +++ b/mysql-test/suite/compat/oracle/r/func_decode.result > @@ -1,8 +1,8 @@ > SET sql_mode=ORACLE; > SELECT DECODE(10); > -ERROR 42000: Incorrect parameter count in the call to native function 'DECODE' > +ERROR 42000: Incorrect parameter count in the call to native function 'oracle_schema.DECODE' Hmm, may be this should say DECODE as before? you know, falls under the case "if you don't change sql_mode back and forth, you won't see schema-qualified names" > SELECT DECODE(10,10); > -ERROR 42000: Incorrect parameter count in the call to native function 'DECODE' > +ERROR 42000: Incorrect parameter count in the call to native function 'oracle_schema.DECODE' > SELECT DECODE(10,10,'x10'); > DECODE(10,10,'x10') > x10 > diff --git a/mysql-test/suite/compat/oracle/r/vcol_innodb.result b/mysql-test/suite/compat/oracle/r/vcol_innodb.result > new file mode 100644 > index 00000000000..9fa97c75c10 > --- /dev/null > +++ b/mysql-test/suite/compat/oracle/r/vcol_innodb.result > @@ -0,0 +1,51 @@ > +SET @table_open_cache=@@GLOBAL.table_open_cache; why do you need to manipulate table_open_cache? to trigger a reopen? Just do flush tables, it's explicit, more readable and more... controllable. > +SET sql_mode=''; > +CREATE TABLE t (d INT,b VARCHAR(1),c CHAR(1),g CHAR(1) GENERATED ALWAYS AS (SUBSTR(b,0,0)) VIRTUAL,PRIMARY KEY(b),KEY g(g)) ENGINE=InnoDB; > +INSERT INTO t VALUES (0); > +ERROR 21S01: Column count doesn't match value count at row 1 > +SET sql_mode='ORACLE'; > +INSERT INTO t SET c=REPEAT (1,0); > +Warnings: > +Warning 1364 Field 'b' doesn't have a default value > +ALTER TABLE t CHANGE COLUMN a b INT; > diff --git a/sql/item_func.h b/sql/item_func.h > index 76a997c33fb..cdbefb82541 100644 > --- a/sql/item_func.h > +++ b/sql/item_func.h > @@ -56,8 +56,40 @@ class Item_func :public Item_func_or_sum, > bool check_argument_types_can_return_date(uint start, uint end) const; > bool check_argument_types_can_return_time(uint start, uint end) const; > void print_cast_temporal(String *str, enum_query_type query_type); > + > + void print_schema_qualified_name(String *to, > + const LEX_CSTRING &schema_name, > + const char *function_name) const I don't see why you'd need this helper. is it something that was used in earlier versions of the patch? > + { > + // e.g. oracle_schema.func() > + to->append(schema_name); > + to->append('.'); > + to->append(function_name); > + } > + > + void print_sql_mode_qualified_name(String *to, > + enum_query_type query_type, > + const char *function_name) const > + { > + const Schema *func_schema= schema(); > + if (!func_schema || func_schema == Schema::find_implied(current_thd)) > + to->append(function_name); > + else > + print_schema_qualified_name(to, func_schema->name(), function_name); > + } > + > + void print_sql_mode_qualified_name(String *to, enum_query_type query_type) > + const > + { > + return print_sql_mode_qualified_name(to, query_type, func_name()); > + } I don't see why you need this helper either, you never use print_sql_mode_qualified_name with the last argument being not func_name(). So you can remove this helper and the third argument of print_sql_mode_qualified_name. > + > public: > > + // Print an error message for a builtin-schema qualified function call > + static void wrong_param_count_error(const LEX_CSTRING &schema_name, > + const LEX_CSTRING &func_name); > + > table_map not_null_tables_cache; > > enum Functype { UNKNOWN_FUNC,EQ_FUNC,EQUAL_FUNC,NE_FUNC,LT_FUNC,LE_FUNC, > diff --git a/sql/item_strfunc.cc b/sql/item_strfunc.cc > index ae078dbb22f..92d5e196da4 100644 > --- a/sql/item_strfunc.cc > +++ b/sql/item_strfunc.cc > @@ -2170,13 +2170,31 @@ bool Item_func_trim::fix_length_and_dec() > > void Item_func_trim::print(String *str, enum_query_type query_type) > { > + LEX_CSTRING suffix= {STRING_WITH_LEN("_oracle")}; > if (arg_count == 1) > { > - Item_func::print(str, query_type); > + if (query_type & QT_FOR_FRM) > + { > + // 10.3 downgrade compatibility for FRM > + str->append(func_name()); > + if (schema() == &oracle_schema_ref) > + str->append(suffix); > + } > + else > + print_sql_mode_qualified_name(str, query_type, func_name()); > + print_args_parenthesized(str, query_type); > return; > } > - str->append(Item_func_trim::func_name()); > - str->append(func_name_ext()); > + > + if (query_type & QT_FOR_FRM) > + { > + // 10.3 downgrade compatibility for FRM > + str->append(Item_func_trim::func_name()); > + if (schema() == &oracle_schema_ref) > + str->append(suffix); > + } > + else > + print_sql_mode_qualified_name(str, query_type, Item_func_trim::func_name()); it'd be simpler if you move the above block that prints the function name before if (arg_count == 1) also you won't need suffix, but can do like in all other functions str->append(STRING_WITH_LEN("trim_oracle"); > str->append('('); > str->append(mode_name()); > str->append(' '); > diff --git a/sql/sql_lex.cc b/sql/sql_lex.cc > index 71f592a3852..bb53d1a510a 100644 > --- a/sql/sql_lex.cc > +++ b/sql/sql_lex.cc > @@ -2084,7 +2100,64 @@ bool Lex_input_stream::get_7bit_or_8bit_ident(THD *thd, uchar *last_char) > } > > > -int Lex_input_stream::scan_ident_sysvar(THD *thd, Lex_ident_cli_st *str) > +/* > + Resolve special SQL functions that have a qualified syntax in sql_yacc.yy. > + These functions are not listed in the native function registry > + because of a special syntax, or a reserved keyword: > + > + mariadb_schema.SUBSTRING('a' FROM 1 FOR 2) -- Special syntax I didn't find it in Oracle's manual, by the way > + mariadb_schema.TRIM(BOTH ' ' FROM 'a') -- Special syntax > + mariadb_schema.REPLACE('a','b','c') -- Verb keyword > +*/ > + > +int Lex_input_stream::find_keyword_qualified_special_func(Lex_ident_cli_st *str, > + uint length) const > +{ > + /* > + There are many other special functions, see the following grammar rules: > + function_call_keyword > + function_call_nonkeyword > + Here we resolve only those that have a qualified syntax to handle > + different behavior in different @@sql_mode settings. > + > + Other special functions do not work in qualified context: > + SELECT mariadb_schema.year(now()); -- Function year is not defined > + SELECT mariadb_schema.now(); -- Function now is not defined > + > + We don't resolve TRIM_ORACLE here, because it does not have > + a qualified syntax yet. Search for "trim_operands" in sql_yacc.yy > + to find more comments. > + */ > diff --git a/sql/sql_schema.h b/sql/sql_schema.h > index 1174bc7a83f..2c52646f2ea 100644 > --- a/sql/sql_schema.h > +++ b/sql/sql_schema.h > @@ -77,5 +98,6 @@ class Schema > > > extern Schema mariadb_schema; > +extern const Schema &oracle_schema_ref; What's the difference between these two definitions. Do you expect someone will need to change mariadb_schema? > > #endif // SQL_SCHEMA_H_INCLUDED Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 2

Eliminating sprintf deprecation warnings (maybe others as well)
by Stein Vidar Hagfors Haugan 27 Oct '23

27 Oct '23

Hi! First time here, so my apologies if I'm violating some spoken or unspoken rules :) I'm developing/maintaining two specialized storage engines(*) and I have one issue: There are numerous uses of sprintf() in the general mariadb codebase, and each one generates a warning during compilation when doing a plain cmake/make build (on MacOS & RHEL7 at least). I am a bit fanatical about hunting down and eliminating warnings if at all possible, no matter how small and insignificant - if you become accustomed to routinely getting warnings during compilation, you can easily overlook an important one. So I am considering to contribute with a patch that will eliminate all sprintf() instances by converting them to snprintf(destination, N, ...). Of course, hunting down the correct N for each instance can be a *lot* of work (I presume this is why this hasn't been done already). So I'm proposing to make a macro UNSAFE_SNPRINTF(destination, ...) which would expand to snprintf(destination, 10000, ...) and replace all current sprintf() calls with that. Yes, 10000 is a ridiculously high number that makes it look like we're introducing a huge security risk, but but the current sprintf() calls have an implicit N = infinity. In some instances, the compiler sniffs out the actual size of the buffer and complains - in that case I would change the code to an appropriate regular snprintf() call. Would doing this be worthwile? I.e., is there a good chance that such a patch would actually be accepted into the codebase? If so, I might also do the same with other warnings (I believe there are some type-punning warnings at least). Sincerely, Stein Haugan *) The engines are for internal use so far, but might make them public at some point

2 2

Re: [PATCH 4/4] MDEV-31273: Precompute binlog checksums
by Michael Widenius 26 Oct '23

26 Oct '23

Hi! Review of MDEV-31273: Precompute binlog checksums On Fri, Aug 25, 2023 at 10:16 AM Kristian Nielsen <knielsen(a)knielsen-hq.org> wrote: > > Compute binlog checksums (when enabled) already when writing events > into the statement or transaction caches, where before it was done > when the caches are copied to the real binlog file. This moves the > checksum computation outside of holding LOCK_log, improving > scalabitily. DBUG_RETURN(my_b_copy_to_cache(from_cache, to_cache, SIZE_T_MAX)); You could use from_cache->end_of_file instead of SIZE_T_MAX <cut> uchar checksum_opt; Wouldn't it be better to have this as an "enum_binlog_checksum_alg" to avoid some casts ? > } > else if (data) > return (enum enum_binlog_checksum_alg)data->checksum_opt; > else > return BINLOG_CHECKSUM_ALG_OFF; > } You can remove the else's <cut> ulong *param_ptr_binlog_cache_disk_use, bool precompute_checksums) : stmt_cache(precompute_checksums), trx_cache(precompute_checksums), last_commit_pos_offset(0), using_xa(FALSE), xa_xid(0) { stmt_cache.set_binlog_cache_info(param_max_binlog_stmt_cache_size, param_ptr_binlog_stmt_cache_use, This code was a bit confusing as we first initialize stmt_cache and trx_cache with checksum and then we initialize them again in the next two lines. I understand why and this does not have to be changed, but it does look a bit strange. The other option would be to have set_binlog_cache_info() take checksum as an argument(). <cut> This one confused me a bit: Log_event_writer writer(file, 0, checksum_alg, &crypto); Log_event_writer writer(file, cache_data, checksum_alg, &crypto); Why this change? It may at least affect calls to: void Log_event_writer::add_status(enum_logged_status status) { if (likely(cache_data)) Fix (after discussions on slack): Remove the following lines in MYSQL_BIN_LOG::write_event(Log_event *ev, binlog_cache_data *cache_data, if (cache_data) cache_data->add_status(ev->logged_status()); <cut> > Don't attempt to precompute checksums if: > - Disabled by user request, --binlog-legacy-event-pos > - Binlog is encrypted, cannot use precomputed checksums > - WSREP/Galera. Why Galera? Would be good to explain this in the commit comment. <cut> > /* > If possible, just copy the cache over byte-by-byte with pre-computed > checksums. > */ > if (likely(binlog_checksum_options == cache_data->checksum_opt) && > likely(!crypto.scheme) && > likely(!opt_binlog_legacy_event_pos)) > { > int res= my_b_copy_all_to_cache(cache, &log_file); > status_var_add(thd->status_var.binlog_bytes_written, my_b_tell(cache)); > DBUG_RETURN(res ? ER_ERROR_ON_WRITE : 0); > } I was just wondering if this could sometimes be optimized to write directly to the real binlog file without another cache in between. <cut> bool precomputed_checksums= (cache_data->checksum_opt != BINLOG_CHECKSUM_ALG_OFF); uint old_checksum_len= precomputed_checksums ? BINLOG_CHECKSUM_LEN : 0; Why two variables when one can as easily do: uint old_checksum_len= ((cache_data->checksum_opt != BINLOG_CHECKSUM_ALG_OFF) ? BINLOG_CHECSUM_LEN : 0); In current write_cache() code, group is a bad name. Please rename 'group' to 'log_file_pos'. By the way, great that you removed the 'end_log_file_pos' variable, which also was a very confusing name related how it was used! > /* > Any old precomputed checksum must _not_ be written here. Instead, it > must be discarded; the new checksum, if needed, is written by > writer.write_footer(). > */ > if (ev_len > old_checksum_len) > { > uint bytes_to_skip= > old_checksum_len - std::min(old_checksum_len, ev_len - chunk); > if (writer.write_data(cache->read_pos, chunk - bytes_to_skip)) > goto error_in_write; > } The above is likely wrong as for long events we would execute the inner part multiple times, while there is only on checksum. As checksum is last in the event, why not just do 'even_len-= old_checksum' Before the loop to copy the event and then disregard 'old_checksum_len' bytes from the cache at the end of the loop? Anyway, to find bugs like this, we need to have a test case with events that are bigger than the IO_CACHE size for cache. Setting binlog_file_cache_size to 8192 (min value) should make it easy to test this. <cut> > binlog_checksum_options= value; > my_atomic_storeul_explicit(&binlog_checksum_options, value, MY_MEMORY_ORDER_RELAXED); Atom is not needed as we have a lock on binlog. ...... Something totally different. I noticed in MYSQL_BIN_LOG::write_cache(): group= (size_t)my_b_tell(&log_file); val= uint4korr(header + LOG_POS_OFFSET) + group + end_log_pos_inc; int4store(header + LOG_POS_OFFSET, val); The first log entry is probably ok, as we have done a rotate() event before. However the cache can have a lot of log_events (as part of a transaction). The end_log_pos after the first event can be wrong as it may not fit into 4 bytes. My understanding is that this is not a problem anymore as we are not using end_log_pos anymore. However I still wonder if rotate() should not only consider the current log file size but also the size of all events we plan to write to the log and do a rotate if the total new log file size > 4G. Regards, Monty

2 1

Re: [PATCH 3/4] MDEV-31273: Refactor MYSQL_BIN_LOG::write_cache()
by Michael Widenius 26 Oct '23

26 Oct '23

Hi! On Fri, Aug 25, 2023 at 10:16 AM Kristian Nielsen <knielsen(a)knielsen-hq.org> wrote: > > Preparatory patch for pre-computing binlog checksums outside of holding > LOCK_log. > > The existing code for MYSQL_BIN_LOG::write_cache() was needlessly complex > and very hard to understand and modify for handling the new case where > pre-computed checksums are already present in the IO_CACHE. MDEV-31273: Refactor MYSQL_BIN_LOG::write_cache() > if (likely(length > LOG_EVENT_HEADER_LEN)) > { > header= cache->read_pos; > cache->read_pos+= LOG_EVENT_HEADER_LEN; > length-= LOG_EVENT_HEADER_LEN; > } > else > { > size_t sofar= length; > size_t remain= LOG_EVENT_HEADER_LEN - sofar; > header= &header_buf[0]; > memcpy(header, cache->read_pos, sofar); > cache->read_pos+= sofar; > > while (hdr_offs < length) > length= my_b_fill(cache); > if (!length) > goto error_in_read; > size_t chunk= std::min(length, remain); > memcpy(header + sofar, cache->read_pos, chunk); > sofar+= chunk; > remain-= chunk; > cache->read_pos+= chunk; > length-= chunk; > } while (unlikely(remain > 0)); > } > The above can be replaced with: if (my_b_read(cache, header, LOG_EVENT_HEAD_LENGTH)) goto error_in_read; It is almost exactly as efficent as the above (one extra if) and avoids using cache internals. Note that you do not have to do call my_fill() if you use my_b_read(). my_b_read() will return 0 if it was able to read all data. In case of end of file, my_b_read() will return 1 and info->error will be 0. If needed, you can als find out how much data left to read from IO_CACHE: left_data_to_read= (cache->end_of_file - my_b_tell(cache)) (I can make an inline function of that if needed). /* Write the rest of the event. */ > while (ev_len > 0) > { > if (length == 0) > length= my_b_fill(cache); > if (!length) > goto error_in_read; > -> while (ev_len > 0) { if (length == 0) { if (!(length= my_b_fill(cache))); goto error_in_read; } <cut> uint chunk= std::min(ev_len, (uint)length); I would have prefer to have MY_MIN() used (like the rest of the code). (not critical) Regards, Monty

2 1

Re: ce9ce585e72: MDEV-31184 Remove parser tokens DECODE_MARIADB_SYM and DECODE_ORACLE_SYM
by Sergei Golubchik 23 Oct '23

23 Oct '23

Hi, Alexander, On Oct 21, Alexander Barkov wrote: > revision-id: ce9ce585e72 (mariadb-10.4.30-103-gce9ce585e72) > parent(s): 1fde785315e > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-08-31 13:49:19 +0400 > message: > > MDEV-31184 Remove parser tokens DECODE_MARIADB_SYM and DECODE_ORACLE_SYM > > Changing the code handling sql_mode-dependent function DECODE(): > > - removing parser tokens DECODE_MARIADB_SYM and DECODE_ORACLE_SYM > - removing the DECODE() related code from sql_yacc.yy/sql_yacc_ora.yy > - adding handling of DECODE() with help of a new Create_func_func_decode please, add a test for DECODE_MARIADB_SYM that changes from ER_PARSE_ERROR to ER_WRONG_PARAMCOUNT_TO_NATIVE_FCT. Why sql_yacc.yy had rules for DECODE_ORACLE_SYM and vice versa? Was it possible to get DECODE_ORACLE_SYM in sql_yacc.yy? > diff --git a/mysql-test/suite/compat/oracle/r/func_decode.result b/mysql-test/suite/compat/oracle/r/func_decode.result > index b49bad93627..2809e971be3 100644 > --- a/mysql-test/suite/compat/oracle/r/func_decode.result > +++ b/mysql-test/suite/compat/oracle/r/func_decode.result > @@ -1,8 +1,8 @@ > SET sql_mode=ORACLE; > SELECT DECODE(10); > -ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near ')' at line 1 > +ERROR 42000: Incorrect parameter count in the call to native function 'DECODE' > SELECT DECODE(10,10); > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy > index cdc04d93708..048741b6ca1 100644 > --- a/sql/sql_yacc.yy > +++ b/sql/sql_yacc.yy > @@ -11027,18 +11024,6 @@ function_call_nonkeyword: > if (unlikely($$ == NULL)) > MYSQL_YYABORT; > } > - | DECODE_MARIADB_SYM '(' expr ',' expr ')' > - { > - $$= new (thd->mem_root) Item_func_decode(thd, $3, $5); > - if (unlikely($$ == NULL)) > - MYSQL_YYABORT; > - } > - | DECODE_ORACLE_SYM '(' expr ',' decode_when_list_oracle ')' > - { > - $5->push_front($3, thd->mem_root); > - if (unlikely(!($$= new (thd->mem_root) Item_func_decode_oracle(thd, *$5)))) > - MYSQL_YYABORT; > - } > | EXTRACT_SYM '(' interval FROM expr ')' > { > $$=new (thd->mem_root) Item_extract(thd, $3, $5); > @@ -12209,25 +12194,6 @@ when_list_opt_else: > } > ; > > -decode_when_list_oracle: > - expr ',' expr > - { > - $$= new (thd->mem_root) List<Item>; > - if (unlikely($$ == NULL) || > - unlikely($$->push_back($1, thd->mem_root)) || > - unlikely($$->push_back($3, thd->mem_root))) > - MYSQL_YYABORT; > - > - } > - | decode_when_list_oracle ',' expr > - { > - $$= $1; > - if (unlikely($$->push_back($3, thd->mem_root))) > - MYSQL_YYABORT; > - } > - ; > - > - > /* Equivalent to <table reference> in the SQL:2003 standard. */ > /* Warning - may return NULL in case of incomplete SELECT */ Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 1

Re: 56fa1da9c67: MDEV-30048 Prefix keys for CHAR work differently for MyISAM vs InnoDB
by Sergei Golubchik 23 Oct '23

23 Oct '23

Hi, Alexander, The actual code change (replacing strnncollsp with strnncollsp_nchars) is ok. But I didn't like the functions you've created. The *_ft_* family is quite confusing, I would not expect to see fulltext-specific functions my_compare.h. I would expect to see there functions which are generic and named by what they're doing, not by where they should be used. ha_compare_char_fixed and ha_compare_char_varying are better. Names are clearer and it's kind of understandable what they do. I suggest you remove *_ft_* functions. fulltext code should, I guess, just use charset_info->coll->strnncoll() everywhere. Not strnncollsp(), because words normally don't have trailing spaces, and if some custom pluggable parser returns words with trailing spaces, they're likely a part of the word and should be compared too. Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org On Oct 17, Alexander Barkov wrote: > revision-id: 56fa1da9c67 (mariadb-10.4.28-90-g56fa1da9c67) > parent(s): ed2adc8c6f9 > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-04-07 14:54:17 +0400 > message: > > MDEV-30048 Prefix keys for CHAR work differently for MyISAM vs InnoDB > > Also fixes: MDEV-30050 Inconsistent results of DISTINCT with NOPAD > > Problem: > > Key segments for CHAR columns where compared using strnncollsp() > for engines MyISAM and Aria. > > This did not work correct in case if the engine applyied trailing > space compression. > > Fix: > > Replacing ha_compare_text() with a number of new functions: > > - ha_compare_ft_text_full() > - ha_compare_ft_text_prefix() > - ha_compare_ft_text() > - ha_compare_char_varying() > - ha_compare_char_fixed() > > The code branch corresponding to comparison of CHAR column keys > (HA_KEYTYPE_TEXT segment type) now uses ha_compare_char_fixed() > which calls strnncollsp_nchars(). > > For the rest of the code: > - comparison of VARCHAR/TEXT column keys > (HA_KEYTYPE_VARTEXT1, HA_KEYTYPE_VARTEXT2 segments types) > - comparison in the fulltext code > this patch does not change the behaviour.

2 3

Re: [PATCH] MDEV-31949 parallel slave xa Round-Robin distribution
by Kristian Nielsen 22 Oct '23

22 Oct '23

Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: Hi Andrei, > commit 96bd9e6b780f3738b5008b89aae6b0b086f15943 > Author: Andrei <andrei.elkin(a)mariadb.com> > Date: Sat Aug 19 19:49:25 2023 +0300 > XA-Prepare group of events > > XA START xid > ... > XA END xid > XA PREPARE xid > > and its XA-"complete" terminator > > XA COMMIT or > XA ROLLBACK > > are made distributed Round-Robin across slave parallel workers. > The former hash-based policy was proven to attribute to execution > latency through creating a big - many times larger than the size > of the worker pool - queue of binlog-ordered transactions > to commit. > > Acronyms and notations used below: > > XAP := XA-Prepare event or the whole prepared XA group of events > XAC := XA-"complete", which is a solitary group of events > |W| := the size of the slave worker pool > Subscripts like `_k' denote order in a corresponding sequence > (e.g binlog file). Here is my review of this patch. > The list is arranged as a sliding window with the size of 2*|W| to account > a possibility of XAP_k -> XAP_k+2|W|-1 the largest (in the group-of-events > count sense) dependency. I spent a lot of effort trying to understand why the factor 2 in the size of 2*|W| occurs. Since each transaction must wait_for_prior_commit for the prior transaction, there should never be more than |W| transactions active. In your example with |W|=4, you say that > Worker #4 can take on its T_8 when T_1 is yet at the beginning of its processing But that does not seem possible. T_8 cannot start until T_1 (and 2..4) have done wakeup_subsequent_commits(), which surely shouldn't happen "at the beginning of its processing", but only at the end. But I think I now understand. The problem is that the XA code calls wakeup_subsequent_commits _before_ updating the XID hash (in XA PREPARE) / deleting from the XID hash (in XA COMMIT/ROLLBACK). Right? This I think is the fundamental issue to address. The wait_for_prior_commit mechanism is to ensure that the prior transaction has completed its commit, which means the state of the transaction is "committed" in memory. Thus, the XID hash, which records information about the status of the (XA) transactions, must also be updated before wakeup_prior_commit may be done. I can see that the XA PREPARE/COMMIT/ROLLBACK already uses the trx_group_commit_leader() code. So what you should do is to arrange for the update of / delete from the XID hash happens inside there, just before the wakeup_subsequent_commits() gets called: if (current->cache_mngr->using_xa && likely(!current->error) && DBUG_EVALUATE_IF("skip_commit_ordered", 0, 1)) { mysql_mutex_lock(&current->thd->LOCK_thd_data); run_commit_ordered(current->thd, current->all); mysql_mutex_unlock(&current->thd->LOCK_thd_data); } current->thd->wakeup_subsequent_commits(current->error); I checked the code, the XID hash update/delete currently happens shortly after that in the code, so there should be no problems moving it in there. [Incidentally, I also noticed that the XA PREPARE/COMMIT does not use the commit_ordered mechanism. While this seems (I think) unrelated to this patch, I think it is something you need to look into. The commit_ordered mechansim is central to ensuring that InnoDB and binlog commit in the same order, and I don't see how this is guaranteed for the current XA code (and I suspect this might also mean that binlog recovery will be broken in some cases).] Once the XID hash update/delete is moved as described, we have the very nice property that after an XA PREPARE/COMMIT/ROLLBACK T_i has done wakeup_subsequent_commits(), it is safe for a following T_j that refers to the same XID to replicate. This means that you can now use the existing wait_for_prior_commit mechanism to handle the dependencies between XA event groups with the same XID. Thus, there is no need to introduce a separate (and very complicated) wait mechanism of xid_cache_insert_maybe_wait() and xid_cache_search_maybe_wait(), solely for parallel replication of user XA. Thus, in the SQL driver thread, using your sliding window (which need only be of size |W| I believe), you can mark a "T1: XA COMMIT <xid> " followed by "T2: XA PREPARE <xid>" of a duplicate XID, that T2 must do a wait_for_prior_commit(T1) before running. This is simple, it can use the existing mechanims for that, using entry->last_committed_sub_id. Just like rgi->wait_commit_sub_id and rgi->wait_commit_group_info, we can introduce eg.: rgi->pre_dependency_sub_id rgi->pre_dependency_group_info and then do a wait_for_prior_commit(&rgi->pre_dependency_group_info->commit_orderer) if the rgi->pre_dependency_sub_id > entry->last_committed_sub_id) Similarly, the sliding window can record for "T3: XA PREPARE" and "T4: XA COMMIT", that T4 should do a wait_for_prior_commit(T3) before running. Then T4 will be sure that the XA transaction is ready. This will also be a good preparation for later introducing more of this kind of pre-calculated dependencies to the parallel replication scheduling, along the lines of the "balanced applier" that you have written about previously. These ideas I believe have tremendous potential, and handling this in a generic way is a very big improvement. I'm not suggesting to implement more than what is necessary now for XA, but keep it in mind that this dependency calculation in the SQL driver thread can be used for other than user XA in the future. You can also name the new introduced dependency fields appropriately (ie. with generic names not specific to XA when applicable). This will greatly simplify the patch, I believe; and more importantly it will integrate the XA-required scheduling in a clean and generic way in the parallel replication code. Following is more detailed comments on the patch: > - DBUG_ASSERT( > - !(thd->rgi_slave && > - !thd->rgi_slave->worker_error && > - thd->rgi_slave->did_mark_start_commit) || > - (thd->transaction->xid_state.is_explicit_XA() || > - (thd->rgi_slave->gtid_ev_flags2 & Gtid_log_event::FL_PREPARED_XA))); > - > + DBUG_ASSERT(!(thd->rgi_slave && > + !thd->rgi_slave->worker_error && > + thd->rgi_slave->did_mark_start_commit) || > + (thd->transaction->xid_state.is_explicit_XA() || > + (thd->rgi_slave->gtid_ev_flags2 & > + (Gtid_log_event::FL_PREPARED_XA | > + Gtid_log_event::FL_COMPLETED_XA)))); > if (thd->rgi_slave && > !thd->rgi_slave->worker_error && > thd->rgi_slave->did_mark_start_commit) > thd->rgi_slave->unmark_start_commit(); Why is it allowed to rollback here while did_mark_start_commit is true for XA PREPARE or XA COMMIT? The comment above should be extended to explain this. And why does the following if () statement then still do an emergency unmark_start_commit(), when the condition is not caught by the assertion? It looks like something is wrong in the earlier code, the intention here is that it should never be necessary to unmark_start_commit() here, and if it is, then it is a bug flagged by the assertion. > @@ -1751,7 +1753,8 @@ binlog_flush_cache(THD *thd, binlog_cache_mngr *cache_mngr, > > if ((using_stmt && !cache_mngr->stmt_cache.empty()) || > (using_trx && !cache_mngr->trx_cache.empty()) || > - thd->transaction->xid_state.is_explicit_XA()) > + (thd->transaction->xid_state.is_explicit_XA() || > + (thd->rgi_slave && thd->rgi_slave->is_async_xac))) Why is this extra condition thd->rgi_slave->is_async_xac necessary? There are many of these conditions spread around the code in the patch. I think perhaps it is because the transaction is somehow not "connected" to the THD? Because it is in the XID cache? But the situation must be similar on the master, if the XA COMMIT happens in a different transaction from XA PREPARE. So this should be done the same way on the slave, so that the XA transaction gets connected to the THD of the worker thread processing it, and so that these extra rgi_slave->is_async_xac conditions are not needed. It is very fragile to have such conditions spread around the code, it will be impossible to avoid bugs due to forgetting such extra condition on one place or another. > + /* > + While xid_state.get_xid() is a robust method to access `xid` > + it can't be used on slave by the asynchronously running XA-"complete". > + In the latter case thd->lex->xid is safely accessible. > + */ > + buflen= serialize_with_xid(is_async_xac? thd->lex->xid : > + thd->transaction->xid_state.get_xid(), Same here, it should be possible to use the same xid_state in the slave thread as on the master, so we avoid special cases all over the code for the slave threads. > +static bool acquire_xid(THD *thd) > +{ > + bool rc= false; > + > + if (thd->rgi_slave && thd->rgi_slave->is_async_xac && > + thd->rgi_slave->gtid_ev_flags2 & Gtid_log_event::FL_COMPLETED_XA) > + { > + XID_STATE &xid_state= thd->transaction->xid_state; This function actually seems to do part of "XA transaction gets connected to the THD". But it's called from binlog_commit_by_xid() and binlog_rollback_by_xid(), which means it has to have yet another of these special-condition checks for thd->rgi_slave->is_async_xac etc. Instead, call this from the code that applies the "XA COMMIT" event (eg. in log_event_server.cc), and make sure it fully connects the XA transaction to the the THD. Then you get rid of the checks for is_async_xac in a lot of places, and also avoid polluting with replication details the code than runs the original transaction on the master. > + rc= binlog_commit(thd, TRUE, FALSE); > + thd->ha_data[binlog_hton->slot].ha_info[1].reset(); > + } > + if (!rc) > + { > + rc= acquire_xid(thd); > + } > + if (thd->is_current_stmt_binlog_disabled()) > + { > + thd->wakeup_subsequent_commits(rc); > + } So IIUC, here we first binlog the XA COMMIT and then wakeup_subsequent_commits(). But the engine commit only happens afterwards, in ha_commit_or_rollback_by_xid. Why wakeup_subsequent_commits() here? It seems too early, the transaction is not yet committed in the engine, how do you avoid that the commits in the engine will happen in the wrong order? And then a mariabackup might take a backup with T2 committed but T1 not and not have a valid replication position to provision a slave. And also, is this skipping the binlog transaction coordinator and two-phase commit with the engines? And in particular, commit_ordered()? Then again, it seems we have the problem that the engine can commit in the opposite order from the binlog. I wonder if this will also affect crash recovery if we crash with different commit order in binlog and engine? So I think wakeup_subsequent_commits() here is wrong, should be removed. > @@ -2185,7 +2293,9 @@ int binlog_commit(THD *thd, bool all, bool ro_1pc) > } > > if (cache_mngr->trx_cache.empty() && > - (thd->transaction->xid_state.get_state_code() != XA_PREPARED || > + ((thd->transaction->xid_state.get_state_code() != XA_PREPARED && > + !(thd->rgi_slave && thd->rgi_slave->is_parallel_exec && > + thd->lex->sql_command == SQLCOM_XA_COMMIT)) || As explained above, this condition (and the similar in binlog_rollback() should be avoided. > @@ -10510,13 +10624,20 @@ int TC_LOG_BINLOG::unlog_xa_prepare(THD *thd, bool all) > > binlog_cache_mngr *cache_mngr= thd->binlog_setup_trx_data(); > int cookie= 0; > + int rc= 0; > + > + if (thd->rgi_slave && thd->is_current_stmt_binlog_disabled()) > + { > + rc= thd->wait_for_prior_commit(); > + if (rc == 0) > + thd->wakeup_subsequent_commits(rc); > + return rc; > + } Why is this necessary? If it is really necessary, then put it in the code that applies the XA PREPARE event, not as a random special case in generic code. > - bool rc= false; > - > > rc= write_empty_xa_prepare(thd, cache_mngr); // normally gains need_unlog > > static bool write_empty_xa_prepare(THD *thd, binlog_cache_mngr *cache_mngr) > { > return binlog_commit_flush_xa_prepare(thd, true, cache_mngr); > } Smaller point: since you're fixing the type of `rc` to be int (which is correct), also fix the return type of write_empty_xa_prepare() to be int and not bool - as binlog_commit_flush_xa_prepare() as well as unlog_xa_prepare() return int, not bool. > @@ -3314,16 +3314,22 @@ Gtid_log_event::Gtid_log_event(THD *thd_arg, uint64 seq_no_arg, > XID_STATE &xid_state= thd->transaction->xid_state; > if (is_transactional) > { > - if (xid_state.is_explicit_XA() && > - (thd->lex->sql_command == SQLCOM_XA_PREPARE || > - xid_state.get_state_code() == XA_PREPARED)) > + bool is_async_xac= false; > + if ((xid_state.is_explicit_XA() && > + (thd->lex->sql_command == SQLCOM_XA_PREPARE || > + xid_state.get_state_code() == XA_PREPARED)) || > + (is_async_xac= (thd->rgi_slave && thd->rgi_slave->is_async_xac))) Again, this is the generic GTID event contructor, it shouldn't need this kind of logic. It's very strange that a simple class constructor returns something different depending on which thread it runs in! Hopefully this is no longer necessary after XA transaction gets properly connected to the THD. But else, it should be handled by passing in suitable flag from the caller, or by the caller setting the required modifications after construction. > +using std::max; Please don't. Using std::max() explicit is not long and makes it explicit what `max` implementation is used. > @@ -760,7 +762,8 @@ convert_kill_to_deadlock_error(rpl_group_info *rgi) > return; > err_code= thd->get_stmt_da()->sql_errno(); > if ((rgi->speculation == rpl_group_info::SPECULATE_OPTIMISTIC && > - err_code != ER_PRIOR_COMMIT_FAILED) || > + (err_code != ER_PRIOR_COMMIT_FAILED && > + err_code != ER_XAER_NOTA)) || Why? What is special about the ER_XAER_NOTA error? Is it really safe to speculatively run the XA COMMIT in parallel, and then retry it for any other error than ER_XAER_NOTA ? > @@ -2467,6 +2463,7 @@ free_rpl_parallel_entry(void *element) > } > mysql_cond_destroy(&e->COND_parallel_entry); > mysql_mutex_destroy(&e->LOCK_parallel_entry); > + e->concurrent_xaps_window.~Dynamic_array(); > my_free(e); No, let's not do this, calling a destructor explicitly in a POD managed by my_malloc() / my_free(). Why not just make concurrent_xaps_window a pointer to the Dynamic_array and manage with new / delete? Then you can also allocate it lazily, so it will not need to be allocated at all for the 99.9xx% of users who are not replicating user XA transactions. The size of the sliding window is fixed anyway (until slave_parallel_threads is changed), so why use Dynamic_array at all, why not just allocate a plain array of the right size? > +template <> > +struct std::hash<XID> > +{ > + std::size_t operator()(const XID& xid) const > + { > + return my_hash_sort(&my_charset_bin, xid.key(), xid.key_length()); I don't understand, what is the purpose of introducing this function object? Why not just call my_hash_sort() directly (seems it's only used in one place) and avoid this code-obfuscation? > + Dynamic_array<std::pair<std::size_t, uint32>> concurrent_xaps_window; This is dangerous as Dynamic_array has a destructor but rpl_parallel_entry is a POD where we don't call any constructor/destructor. Should instead use a pointer to an object constructed with new (or not use Dynamic_array at all as suggested above). > + /* > + When true indicates that the user xa transaction is going to > + complete (with COMMIT or ROLLBACK) by the worker thread, > + *while* another worker is still preparing it. Once the latter is done > + the xid will be acquired and the flag gets reset. > + */ > + bool is_async_xac; > + I don't understand this comment. Surely it is not possible to *complete*, with XA COMMIT, a transaction before the corresponding XA PREPARE has completed (possibly in another worker thread)? Do you mean that it is possible for a worker thread W2 to *start* applying the XA COMMIT speculatively, before another worker W1 has completed the XA PREPARE? But I think this is not necessary, see below the discussion that the XA COMMIT worker can just do a wait_for_prior_commit() on the event group of its XA PREPARE. Then this should not be necessary. > @@ -137,7 +137,7 @@ template <class Elem> class Dynamic_array > */ > Elem& at(size_t idx) > { > - DBUG_ASSERT(idx < array.elements); > + DBUG_ASSERT(idx < max_size()); > return *(((Elem*)array.buffer) + idx); No, this cannot possibly be right. This is a patch about XA replication, why would it suddenly allow accessing deleted elements of any Dynamic_array used in the server?!? Surely you can allocate the elements you need for your sliding window. In fact, since the window is fixed size anyway, why use Dynamic_array at all? @@ -2367,6 +2367,15 @@ struct wait_for_commit event group is fully done. */ bool wakeup_blocked; + /* + The condition variable servers as a part of facilities to handle various + commit time additional dependency between groups of replication events, e.g + XA-Prepare -> XA-Commit, or XA-Prepare -> XA-Prepare all with the same xid. + */ + mysql_cond_t COND_wait_commit_dep; No, this doesn't belong in struct wait_for_commit. wait_for_commit is a low-level mechanism for ordering commits, it should not need any knowledge even of replication, certainly not of user-XA replication. The COND_wait_commit_dep isn't even used in any function related to wait_for_commit or even sql_class.cc, it's only used in sql/xa.cc. [But as discussed at the start, we can use the existing wait_for_prior_commit instead of introducing this new mechanism just for replicating user XA.] > -static XID_cache_element *xid_cache_search(THD *thd, XID *xid) > +XID_cache_element *xid_cache_search(THD *thd, XID *xid) Why remove the `static`? The function is not used outside of sql/xa.cc > @@ -254,16 +259,221 @@ static XID_cache_element *xid_cache_search(THD *thd, XID *xid) > + if (thd->rgi_slave && thd->rgi_slave->is_parallel_exec) > + { > + DBUG_ASSERT(thd->lex->sql_command == SQLCOM_XA_COMMIT || > + thd->lex->sql_command == SQLCOM_XA_ROLLBACK); > + thd->rgi_slave->is_async_xac= true; Here again we have some logic that relates to replication being inserted somewhat randomly in low-level code, requiring this thd->rgi_slave->is_parallel_exec condition. Instead of doing this, the caller higher up should handle this depending on the return of xid_cache_search(). This way, code paths that are not related to replication don't get affected by the replication-specific logic. > +bool xid_cache_insert_maybe_wait(THD* thd) > +{ This new wait mechanism just for parallel replication of user XA seems very complex. Spin loop, and extra locking and condition variables and locking on a lock-free hash. As explained at the start, let's instead avoid introducing a new mechanism and use the existing wait_for_prior_commit, which is central to parallel replication and highly optimized for this. IIUC, the real problem here is that after wakeup_subsequent_commits() of XA COMMIT, the XID is still in the xid_hash. This is because of this code in trans_xa_commit(): ha_commit_or_rollback_by_xid(thd->lex->xid, !res); if (!res && thd->is_error()) { // hton completion error retains xs/xid in the cache, // unless there had been already one as reflected by `res`. res= true; goto _end_external_xid; } xid_cache_delete(thd, xs); xid_deleted= true; ha_commit_or_rollback_by_xid() ends up in queue_for_group_commit() which will do the wakeup_subsequent_commits. But the xid_cache_delete() happens just after that. This is too late. The wakeup_subsequent_commits() must not occur until after the transaction is fully committed in memory. So instead, move the xid_cache_delete() call so it happens inside the group commit code. Then when wakeup_subsequent_commits() is done after XA COMMIT, the XID entry will be gone. It would go here in MYSQL_BIN_LOG::trx_group_commit_leader(): if (current->cache_mngr->using_xa && likely(!current->error) && DBUG_EVALUATE_IF("skip_commit_ordered", 0, 1)) { mysql_mutex_lock(&current->thd->LOCK_thd_data); run_commit_ordered(current->thd, current->all); mysql_mutex_unlock(&current->thd->LOCK_thd_data); } current->thd->wakeup_subsequent_commits(current->error); I noticed that commit_ordered is not done for user XA transactions in replication. The direct reason for that seems to be that they do not use log_and_order(), which sets cache_mngr->using_xa. Can you explain why the XA PREPARE and XA COMMIT is not following the commit_ordered protocol, and what mechanism exists instead to ensure that everything still works correctly? For example that commit order in binlog and InnoDB will be the same, and that binlog recovery works? And finally, just for the record, I still do not agree with the way replication of user XA has been changed (aka MDEV-32020), and this patch does not fix the underlying problem. But changing it to use the wait_for_prior_commit() mechanism as explained at the start will improve the integration with the parallel replication code and be a step towards solving also some of the other underlying issues. - Kristian.

1 0

Locking thd->LOCK_thd_data during group commit
by Kristian Nielsen 21 Oct '23

21 Oct '23

Hi Serg, I came upon this commit by you: > commit 6b685ea7b0776430d45b095cb4be3ef0739a3c04 > Author: Sergei Golubchik <serg(a)mariadb.org> > Date: Wed Sep 28 18:55:15 2022 +0200 > > correctness assert > > thd_get_ha_data() can be used without a lock, but only from the > current thd thread, when calling from anoher thread it *must* > be protected by thd->LOCK_thd_data > > * fix group commit code to take thd->LOCK_thd_data > * remove innobase_close_connection() from the innodb background thread, > it's not needed after 87775402cd0c and was failing the assert with > current_thd==0 > @@ -8512,7 +8512,11 @@ MYSQL_BIN_LOG::trx_group_commit_leader(group_commit_entry *leader) > ++num_commits; > if (current->cache_mngr->using_xa && likely(!current->error) && > DBUG_EVALUATE_IF("skip_commit_ordered", 0, 1)) > + { > + mysql_mutex_lock(&current->thd->LOCK_thd_data); > run_commit_ordered(current->thd, current->all); > + mysql_mutex_unlock(&current->thd->LOCK_thd_data); > + } This seems _really_ expensive :-( This code runs for every transaction commit in replication, and it runs under a global LOCK_commit_ordered which needs to be held for as short as possible. The commit message doesn't mention anything about what goes wrong during the group commit code. And the patch doesn't have any test case that shows the problem, it just adds an assertion that will trigger during the group commit. So what is the actual reason for this change? Note that at this point, the group commit leader thread _is_ effectively acting as if it was each of the participating threads, the other threads are blocked from running during commit_ordered. If it is just so that we can have this assertion, then that needs to be rolled back. We _really_ don't want to take/release N mutexes in release builds during LOCK_commit_ordered, just to have an assertion in debug builds. If there is some reason, then the correct fix could be to set current_thd for the duration of run_commit_ordered, which will satisfy the assertion and presumably the actual bug? Well, it depends _what_ the actual reason/bug is. - Kristian.

3 5

Re: Review of MDEV-31273: Eliminate Log_event::checksum_alg
by Kristian Nielsen 17 Oct '23

17 Oct '23

Hi Monty, Lots of good comments, thanks! Replies inline: Michael Widenius <michael.widenius(a)gmail.com> writes: > enum enum_binlog_checksum_alg > > In c++ we can drop the first 'enum' > You can fix that by doing: Ack, will fix. > event.select_checksum_alg() > > This is called a lot. Would it not be easier to store the checksum_alg > in the event when it is created and access the variable directly? > On the other hand, the number of calls to event.select_checksum_alg() may > not change as we will only create one Yes, this is called only once per event, when writing the event into the binlog or the event cache. So it seems unnecessary to use extra memory to store it. And we don't necessarily even have binlog_cache_data when constructing the event to pass to select_checksum_alg(), so I think it is correct to select the checksum algorithm to use only when we write the event. > @@ -1258,11 +1258,10 @@ Log_event* Log_event::read_log_event(const > uchar *buf, uint event_len, > > if (ev) > { > - ev->checksum_alg= alg; > #ifdef MYSQL_CLIENT > if (ev->checksum_alg != BINLOG_CHECKSUM_ALG_OFF && > ev->checksum_alg != BINLOG_CHECKSUM_ALG_UNDEF) > ev->crc= uint4korr(buf + (event_len)); > + ev->read_checksum_alg= alg; > > Why move the setting of read_checksum_alg inside MYQL_CLIENT? It is because with this patch, ev->checksum_alg is no longer needed in the server (the class member is removed). It is needed in the client (mysqlbinlog), so I added a new member read_checksum_alg, which is the algorithm that was used to read the event. This is needed so that mysqlbinlog can output the checksum (if any) in textual form. But it is not needed in the server > To ensure that this is not wrongly used, please add > #ifdef MYSQL_CLIENT > > here: > > class Log_event: > > #ifdef MYSQL_CLIENT > + enum enum_binlog_checksum_alg read_checksum_alg; > #endif Yes, agree, that is the intention - the read_checksum_alg is only present in the class in the client, not in the server (and we want to avoid using memory for it in the server). This code is already in the #else branch of #ifdef MYSQL_SERVER: #ifdef MYSQL_SERVER ... #else /* The checksum algorithm used (if any) when the event was read. */ enum enum_binlog_checksum_alg read_checksum_alg; I thought this would have the same effect as your suggestion with #ifdef MYSQL_CLIENT. It probably does, but it seems somewhat complex when MYSQL_CLIENT and MYSQL_SERVER are defined. Maybe they can even both be defined, though probably not where log_event.h is included? It's a bit confusing that we have both MYSQL_CLIENT and MYSQL_SERVER used for conditional compilation in log_event.h. But probably not something that should be cleaned up in this patch. Do you think I should still add the #ifdef MYSQL_CLIENT as you suggested (even though it should be redundant in the #else branch of #ifdef MYSQL_SERVER)? It can't hurt after all, even if we always have MYSQL_CLIENT == !MYSQL_SERVER. > in Format_description_log_event(const uchar > > We have: > > else > { > checksum_alg= BINLOG_CHECKSUM_ALG_UNDEF; > used_checksum_alg= BINLOG_CHECKSUM_ALG_OFF; > } > > The used_checksum_alg is not needed as you set it at function start. Sorry, I don't understand? At the start we set it to "undef": + used_checksum_alg= BINLOG_CHECKSUM_ALG_UNDEF; This value will remain if there's an error return; else the else branch will change it to "off". Did you miss that it's set differently at the start of the function and in this else { } branch? Or did I miss something? > sql/log_event.h > > enum enum_binlog_checksum_alg set_checksum_alg(enum > enum_binlog_checksum_alg alg) > As we are just copying and restoring checksum_length, why not just do > > Format_description_log_event::write(Log_event_writer *writer) > > org_checksum_length= checksum_length... > .... > checksum_length= org_checksum_length; > > instead of calling set_checksum_alg() again ? > (The last set we can even do undonditionally which saves us an if) Agree, that code looks silly, I don't know why I did it that way. I'll fix it, remove the function set_checksum_alg() and just do: org_checksum_length= writer->checksum_length; writer->checksum_length= BINLOG_CHECKSUM_LEN; ... writer->checksum_length= org_checksum_length; > log_event_server.cc > > enum enum_binlog_checksum_alg Log_event::select_checksum_alg() > > Much simpler, but we do not check if cache_type is correct > > DBUG_ASSERT(((get_type_code() != ROTATE_EVENT && > get_type_code() != STOP_EVENT) || > get_type_code() != FORMAT_DESCRIPTION_EVENT) || > cache_type == Log_event::EVENT_NO_CACHE); > > Do you think this is not needed? > (This could be added just before: > > return BINLOG_CHECKSUM_ALG_OFF; Agree, I was happy to remove a lot of complexity and the need for a number of asserts, but this assertion could be kept. I'll add it back. > Anyway, please remove the else after: > > if (cache_type == Log_event::EVENT_NO_CACHE) > return (enum_binlog_checksum_alg)binlog_checksum_options; Ack will do. > Format_description_log_event::write(Log_event_writer *writer) > > > uint8 checksum_byte= (uint8) (used_checksum_alg != BINLOG_CHECKSUM_ALG_UNDEF ? > used_checksum_alg : BINLOG_CHECKSUM_ALG_OFF); > > Is it possible the alg can be UNDEF? > You have several asserts in the code that this should not happen. Yeah, honestly, I'm not sure. It's one of the annoying things in the code that I did not get to fix, that we have both the UNDEF and the OFF state for the checksum algorithm. Format_description code is not performance critical (only once per binlog file), so I suggest I leave the extra check for UNDEF as defensive code, but add an assertion that it is not UNDEF and see if it passes the testsuite. Does that sound ok? > slave.cc > > @@ -1870,10 +1873,8 @@ static int get_master_version_and_clock(MYSQL* > mysql, Master_info* mi) > until it has received a new FD_m. > */ > mi->rli.relay_log.description_event_for_queue->checksum_alg= > mi->rli.relay_log.relay_log_checksum_alg; > > Not sure why this is removed. > > Are these guaranteed to be the same here? > If yes, why not add an assert for this? This patch removes Log_event::checksum_alg and instead adds a field Format_description_log_event::used_checksum_alg which gets set in the Format_description_log_event constructor. So this assignment - mi->rli.relay_log.description_event_for_queue->checksum_alg= - mi->rli.relay_log.relay_log_checksum_alg; is replaced by this code just above which serves the same purpose: - Format_description_log_event(4, mysql->server_version); + Format_description_log_event(4, mysql->server_version, + mi->rli.relay_log.relay_log_checksum_alg); So I think adding an assert DBUG_ASSERT(mi->rli.relay_log.description_event_for_queue->checksum_alg == mi->rli.relay_log.relay_log_checksum_alg); is somewhat redundant, but let me know if you still think it will be useful and I'll add it. - Kristian.

2 1

Re: [MariaDB commits] [PATCH 1/4] MDEV-31273: Replace Log_event::writer with function parameter
by Kristian Nielsen 16 Oct '23

16 Oct '23

Hi Monty, [Following up on review on IRC] Kristian Nielsen via commits <commits(a)lists.mariadb.org> writes: > Replace Log_event::writer with just passing the writer object as a > function parameter to Log_event::write(). We wondered how the compiled code will look when passing the writer object as a parameter to Log_event::write() vs. temporarily storing it in the Log_event object. Interestingly, the code is actually "better" after my patch despite an extra parameter being passed along. This is because the compiler can cache the function parameter in a register across nested calls. While with a class member, the compiler cannot determine that the member will not be changed by the nested calls, so the value needs to be repeatedly re-loaded from memory. Some more details on this finding below. My reason for the change was probably more general, based on this code: > - ev->writer= this; > - int res= ev->write(); > - IF_DBUG(ev->writer= 0,); // writer must be set before every Log_event::write > + int res= ev->write(this); It's clear that the intention with this code is to pass the Log_event_writer object `this` to the ev->write() method, and it's a strange way to do it to temporarily set it as a member in the object rather than using a function parameter. Needing an IF_DBUG() to enforce/check that this is done correctly is a warning sign, and I was worried that as I did further changes I would introduce subtle bugs this way. It is interesting that this makes the code clearer not only for the developer (me), but also for the compiler in this case. Function parameter passing is quite well optimized on amd64, but it was still good to see this in detail in this case. And as you predicted we learned something from it! As we discussed, I will extend the commit message to explain these reasons better. Thanks for the suggestion; it's always good in the reviews to question the reason for changes like this. To check the compiled code in detail, I did a release build (BUILD/compile-pentium64-max) with and without this patch, and disassembled the User_var_log_event::write() function, which has a number of calls now passing the writer object as a function parameter: > - return write_header(event_length) || > - write_data(buf, sizeof(buf)) || > - write_data(name, name_len) || > - write_data(buf1, buf1_length) || > - write_data(pos, val_len) || > - write_data(&flags, unsigned_len) || > - write_footer(); > + return write_header(writer, event_length) || > + write_data(writer, buf, sizeof(buf)) || > + write_data(writer, name, name_len) || > + write_data(writer, buf1, buf1_length) || > + write_data(writer, pos, val_len) || > + write_data(writer, &flags, unsigned_len) || > + write_footer(writer); It looks like this would need to pass the `writer` argument many times to called functions. However, write_data() and write_footer() are actually inline functions: bool write_data(const uchar *buf, size_t data_length) { return writer->write_data(buf, data_length); } bool write_footer() { return writer->write_footer(); } So it's actually only write_header() where we now pass an extra parameter. For write_data() and write_footer() it was already being passed as the (implicit) `this` parameter. With the patch, the compiler saves the `writer` parameter in register r12 and uses it for each call: movq %rsi, %r12 ... movq %r12, %rsi # `writer` pointer movq %rbx, %rdi # `this` pointer call _ZN9Log_event12write_headerEP16Log_event_writerm@PLT ... movl $4, %edx # `data_length` parameter movq %r12, %rdi # `writer` pointer call _ZN16Log_event_writer10write_dataEPKhm@PLT ... leaq 137(%rbx), %rsi # `buf` parameter movq %r14, %rdx # `data_length` parameter movq %r12, %rdi # `writer` pointer call _ZN16Log_event_writer10write_dataEPKhm@PLT ... movq %r12, %rdi # `writer` pointer call _ZN16Log_event_writer12write_footerEv@PLT Before the patch, the code for User_var_log_event::write() loads this->writer for each call except write_header(): movq %rbx, %rdi # Here we don't pass `writer`, write_header() loads it from Log_event::writer call _ZN9Log_event12write_headerEm@PLT ... movq 88(%rbx), %rdi # this->writer leaq -126(%rbp), %rsi # `buf` parameter movl $4, %edx # `data_length` parameter call _ZN16Log_event_writer10write_dataEPKhm@PLT ... movq 88(%rbx), %rdi # this->writer leaq 153(%rbx), %rsi # `buf` parameter movq %r13, %rdx # `data_length` parameter call _ZN16Log_event_writer10write_dataEPKhm@PLT ... movq 88(%rbx), %rdi # this->writer call _ZN16Log_event_writer12write_footerEv@PLT I think the reason for the repeated loads of this->writer from 88(%rbx) is because the compiler cannot be sure that this->writer will not be modified by calls to write_data(). So it needs to reload the pointer each time. With the patch the pointer is a function parameter which cannot be modified by nested calls, so it can be saved in a register. Thus, we save some memory loads with my patch for each call of write_data() / write_footer(), which should be a bit faster (though perhaps not noticeably so). In general, passing parameters to functions is cheap (on amd64), as they are passed in registers, and register-to-register moves are very fast on high-performance CPUs, while memory loads have a relative high latency, even when they hit the cache. And like we see here, using local variables helps the compiler do alias analysis and cache values in registers, while this is a lot harder for the compiler to do for class members. - Kristian.

1 0

Re: d4c173ae6df: MDEV-27595 Backport SQL service, introduced by MDEV-19275.
by Sergei Golubchik 16 Oct '23

16 Oct '23

Hi, Alexey, On Oct 16, Alexey Botchkov wrote: > revision-id: d4c173ae6df (mariadb-10.4.30-117-gd4c173ae6df) > parent(s): 837b777bfd7 > author: Alexey Botchkov > committer: Alexey Botchkov > timestamp: 2023-09-05 13:18:12 +0400 > message: > > MDEV-27595 Backport SQL service, introduced by MDEV-19275. > > Fixes to make SQL SERVICE working. > > Conflicts: > storage/spider/spd_table.cc ^^^ better remove these two lines. normally git removes them automatically, and there's no spd_table.cc in this commit anyway. Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: 837b777bfd7: MDEV-27595 Backport SQL service, introduced by MDEV-19275.
by Sergei Golubchik 16 Oct '23

16 Oct '23

Hi, Alexey, On Oct 16, Alexey Botchkov wrote: > revision-id: 837b777bfd7 (mariadb-10.4.30-116-g837b777bfd7) > parent(s): a49b9314c16 > author: Alexey Botchkov > committer: Alexey Botchkov > timestamp: 2023-09-05 13:18:12 +0400 > message: > > MDEV-27595 Backport SQL service, introduced by MDEV-19275. > > The SQL SERVICE backported into the 10.4. > > diff --git a/plugin/test_sql_service/CMakeLists.txt b/plugin/test_sql_service/CMakeLists.txt > new file mode 100644 > index 00000000000..615508bdc4e > --- /dev/null > +++ b/plugin/test_sql_service/CMakeLists.txt > @@ -0,0 +1,18 @@ > +# Copyright (C) 2013 Alexey Botchkov and SkySQL Ab > +# > +# This program is free software; you can redistribute it and/or modify > +# it under the terms of the GNU General Public License as published by > +# the Free Software Foundation; version 2 of the License. > +# > +# This program is distributed in the hope that it will be useful, > +# but WITHOUT ANY WARRANTY; without even the implied warranty of > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > +# GNU General Public License for more details. > +# > +# You should have received a copy of the GNU General Public License > +# along with this program; if not, write to the Free Software > +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA > + > +SET(SOURCES test_sql_service.c) > + > +MYSQL_ADD_PLUGIN(test_sql_service ${SOURCES} MODULE_ONLY) COMPONENT Test > diff --git a/plugin/test_sql_service/COPYING b/plugin/test_sql_service/COPYING > new file mode 100644 > index 00000000000..6e475df5526 > --- /dev/null > +++ b/plugin/test_sql_service/COPYING > @@ -0,0 +1,339 @@ > + GNU GENERAL PUBLIC LICENSE No need to, we already have one in the CMAKE_SOURCE_DIR, why do you want to copy it just for one plugin? Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: About Release Model
by Kristian Nielsen 16 Oct '23

16 Oct '23

Hi Sergei, Sergei Golubchik <serg(a)mariadb.org> writes: > You've said that it's unclear how the release model works, when we > release what and where a feature can go. > > I've documented the current release model in the knowledge base, > https://mariadb.com/kb/en/mariadb-release-model/ > would you like to take a look and see if it answers your questions? This is a fine overview page, and if this info was not previously available, then it's surely good to have it documented. However, it doesn't help explain how to meaningfully participate as a core developer in the releases. Where does the discussion happen about the progress and planning of the release, and addressing problems? How to partipate in those discussions? How to push a patch into a preview release, or a release, how to make a release? Etc. My assumption is that this is all restricted to employees? But even that isn't explained on the page. - Kristian.

2 3

Question about lacking git commit message in dc9498b
by Otto Kekäläinen 12 Oct '23

12 Oct '23

Hi Sergei and others! While reviewing changes in MariaDB 10.11.5 I was wondering about this commit: https://github.com/mariadb/server/commit/dc9498beb6f79f08edd02fb7f7153e5150… It has only the default message from issuing 'git revert' and a human written explanation *why* the revert was done is missing. AFAIK the reverted commit https://github.com/MariaDB/server/commit/1d0e3d80d84c2b375a444cdbb41c28afd7… was correct and good - so why was it reverted?

2 1

Re: 6cb526bd363: MDEV-32123: require_secure_transport doesn't allow TCP connections
by Sergei Golubchik 10 Oct '23

10 Oct '23

Hi, Dmitry, Looks pretty good! Basically no comments, some minor nitpicking below, feel free to ignore. On Oct 10, Dmitry Shulga wrote: > revision-id: 6cb526bd363 (mariadb-11.2.1-9-g6cb526bd363) > parent(s): 872ed5342d8 > author: Dmitry Shulga > committer: Dmitry Shulga > timestamp: 2023-10-03 18:07:35 +0700 > message: > > MDEV-32123: require_secure_transport doesn't allow TCP connections > > In case the option require_secure_transport is on the user can't > establish a secure ssl connection over TCP protocol. Inability to set up > a ssl session over TCP was caused by the fact that a type of client's > connection was checked before ssl handshake performed (ssl handshake > happens at the function acl_authenticate()). At that moment vio type has > the value VIO_TYPE_TCPIP for client connection that uses TCP transport. > In result, checking for allowable vio type for fails despite the fact > that SSL session being established. To fix the issue move checking of > vio type for allowable values inside the function > parse_client_handshake_packet() > right after client's capabilities discovered that SSL is not requested > by the client. > > diff --git a/mysql-test/main/require_secure_transport.result b/mysql-test/main/require_secure_transport.result > index 94474d0fa0e..8fa273cbae4 100644 > --- a/mysql-test/main/require_secure_transport.result > +++ b/mysql-test/main/require_secure_transport.result > @@ -1,8 +1,13 @@ > -CREATE TABLE t1 (t int(1)); > SET GLOBAL require_secure_transport=ON; > -ERROR HY000: Connections using insecure transport are prohibited while --require_secure_transport=ON. > +ERROR 08004: Connections using insecure transport are prohibited while --require_secure_transport=ON. I wonder why sqlstate has changed now, as far as I remember it was the previous patch that changed it. Was the test skipped until now? > +SELECT (VARIABLE_VALUE <> '') AS have_ssl FROM INFORMATION_SCHEMA.SESSION_STATUS WHERE VARIABLE_NAME='Ssl_cipher'; > +have_ssl > +1 > +disconnect with_ssl; > connection default; > SET GLOBAL require_secure_transport=OFF; > +SELECT (VARIABLE_VALUE <> '') AS have_ssl FROM INFORMATION_SCHEMA.SESSION_STATUS WHERE VARIABLE_NAME='Ssl_cipher'; > +have_ssl > +0 > disconnect without_ssl; > connection default; > -DROP TABLE t1; > diff --git a/mysql-test/main/require_secure_transport.test b/mysql-test/main/require_secure_transport.test > index e238e732423..f3da73a5167 100644 > --- a/mysql-test/main/require_secure_transport.test > +++ b/mysql-test/main/require_secure_transport.test > @@ -1,15 +1,19 @@ > -- source include/have_ssl_communication.inc > -CREATE TABLE t1 (t int(1)); > SET GLOBAL require_secure_transport=ON; > --disable_query_log > --error ER_SECURE_TRANSPORT_REQUIRED > connect without_ssl,localhost,root,,,,,TCP NOSSL; > + > +connect with_ssl,localhost,root,,,,,TCP SSL; > --enable_query_log I'd suggest to remove disable/enable_query_log and let connect commmand show in the result file. It'll make them easier to read. You'll need replace_result before the failing connect to replace the port away. > +SELECT (VARIABLE_VALUE <> '') AS have_ssl FROM INFORMATION_SCHEMA.SESSION_STATUS WHERE VARIABLE_NAME='Ssl_cipher'; > +disconnect with_ssl; > + > connection default; > SET GLOBAL require_secure_transport=OFF; > --disable_query_log > connect without_ssl,localhost,root,,,,,TCP NOSSL; > --enable_query_log > +SELECT (VARIABLE_VALUE <> '') AS have_ssl FROM INFORMATION_SCHEMA.SESSION_STATUS WHERE VARIABLE_NAME='Ssl_cipher'; > disconnect without_ssl; > connection default; > -DROP TABLE t1; Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: 217a7a7ceff: online alter: rework savepoints
by Sergei Golubchik 09 Oct '23

09 Oct '23

Hi, Nikita, On Sep 29, Nikita Malyavin wrote: > revision-id: 217a7a7ceff (mariadb-11.2.1-10-g217a7a7ceff) > parent(s): ad6a305d84d > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-09-23 11:46:36 +0400 > message: > > online alter: rework savepoints > > Use standard handlerton functions for savepoint add/rollback. > > To identify the savepoint, the pointer passed is used. > > Every table that has online alter in progress maintains a list of > savepoints independently. > > Also this removes setting a value to a global variable savepoint_alloc_size > without any protection, which was a race condition bug. > > --- a/sql/online_alter.cc > +++ b/sql/online_alter.cc > @@ -6,7 +6,19 @@ > > static handlerton *online_alter_hton; > > -using Online_alter_cache_list= ilist<online_alter_cache_data>; > +using sv_id_t= void *; typedef, please > + > +struct Online_alter_cache_list: ilist<online_alter_cache_data> > +{ > + sv_id_t savepoint_id= 0; > +}; > @@ -281,13 +308,11 @@ static int online_alter_log_init(void *p) > { > online_alter_hton= (handlerton *)p; > online_alter_hton->db_type= DB_TYPE_ONLINE_ALTER; > - online_alter_hton->savepoint_offset= sizeof(my_off_t); > + online_alter_hton->savepoint_offset= 1; you can even use 0, I guess. Otherwise it'll still be 8, as the server will need to align allocation, I suppose. But I suggest to try to swap the implementation. Make online_alter_hton->savepoint_offset=sizeof(void*); and in savepoint_set create a list or an array of all offsets and store it in the savepoint. Something like this: struct { uint savepoints; my_off_t log_prev_pos[1]; } oasv; int online_alter_savepoint_set(handlerton *hton, THD *thd, oasv **sv) { uint count= 0, i=0; for (auto &cache: cache_list) count++; DBUG_ASSERT(count); *sv= malloc(sizeof(oasv) + sizeof(my_off_t)*(count - 1)); (*sv)->savepoints= count; for (auto &cache: cache_list) (*sv)->log_prev_pos[i]= cache.get_byte_position(); } and so on. You just need to make sure you always use push_back and not push_front for adding caches to the cache_list. instead of a structure with a counter you can use an array with MY_OFF_T_UNDEF (not 0!) at the end. > online_alter_hton->close_connection= online_alter_close_connection; > > - online_alter_hton->savepoint_set= // Done by online_alter_savepoint_set > - [](handlerton *, THD *, void *){ return 0; }; > - online_alter_hton->savepoint_rollback= // Done by online_alter_savepoint_rollback > - [](handlerton *, THD *, void *){ return 0; }; > + online_alter_hton->savepoint_set= online_alter_savepoint_set; > + online_alter_hton->savepoint_rollback= online_alter_savepoint_rollback; > online_alter_hton->savepoint_rollback_can_release_mdl= > [](handlerton *hton, THD *thd){ return true; }; Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 2

Re: ad6a305d84d: online alter: use thd->ha_data to store cache_list
by Sergei Golubchik 09 Oct '23

09 Oct '23

Hi, Nikita, On Sep 29, Nikita Malyavin wrote: > revision-id: ad6a305d84d (mariadb-11.2.1-9-gad6a305d84d) > parent(s): 132c56f1325 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-09-23 11:46:36 +0400 > message: > > online alter: use thd->ha_data to store cache_list nice > diff --git a/sql/online_alter.cc b/sql/online_alter.cc > index 3255fc06c3e..5ad2156cd21 100644 > --- a/sql/online_alter.cc > +++ b/sql/online_alter.cc > @@ -6,6 +6,7 @@ > > static handlerton *online_alter_hton; > > +using Online_alter_cache_list= ilist<online_alter_cache_data>; please, use typedef for that > class online_alter_cache_data: public Sql_alloc, public ilist_node<>, > public binlog_cache_data > @@ -246,7 +268,11 @@ int online_alter_savepoint_rollback(THD *thd, LEX_CSTRING name) > > static int online_alter_close_connection(handlerton *hton, THD *thd) > { > - DBUG_ASSERT(thd->online_alter_cache_list.empty()); > + Ha_data &ha= thd->ha_data[online_alter_hton->slot]; > + auto *cache_list= (Online_alter_cache_list*)ha.ha_ptr; why not thd_get_ha_data()? and don't forget thd_set_ha_data(thd, online_alter_hton, 0); THD can be reused, I suppose, so let's play safe. > + > + DBUG_ASSERT(!cache_list || cache_list->empty()); > + delete cache_list; > return 0; > } Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

2 3

Re: MariaDB 10.11.5 mtr --big results from amd64/arm64/armhf/ppc64el/riscv64/s390x
by Otto Kekäläinen 09 Oct '23

09 Oct '23

Hi! > > I just wanted to share the summary of mariadb-test-run with --big from > > all available architectures on Launchpad > > (https://launchpad.net/~otto/+archive/ubuntu/mariadb/+builds?build_text=&bui…) > > I have tried to fix type_test_double and maria.maria-purge in 10.6. > With 'tried', I mean that I have not been able to repeat the test, but > I have a clue what is going on. > I will push later today. > Any chance you can try to run bb-10.6-monty on launchpad to test my fixes? I cherry-picked these two commits: https://github.com/MariaDB/server/commit/8e093d7597664290b0b5e0eb401b3a180f… https://github.com/MariaDB/server/commit/630cb99823f8e83545ffa16d60c9ac75cd… .. and pushed to https://launchpad.net/~otto/+archive/ubuntu/mariadb/+builds?build_text=&bui… and indeed both tests are now fixed: maria.maria-purge w1 [ pass ] 2134 type_test.type_test_double w1 [ pass ] 59 (from https://launchpadlibrarian.net/691041067/buildlog_ubuntu-mantic-amd64.maria…) Thanks Monty! Last failing that remains after Monty's and Daniel's fixes is: - unit.conc_connection: CONC-640/MDEV-30885

1 0

MariaDB 10.11.5 mtr --big results from amd64/arm64/armhf/ppc64el/riscv64/s390x
by Otto Kekäläinen 07 Oct '23

07 Oct '23

Hi! I just wanted to share the summary of mariadb-test-run with --big from all available architectures on Launchpad (https://launchpad.net/~otto/+archive/ubuntu/mariadb/+builds?build_text=3D&= build_state=3Dall): s390x Completed: Failed 13/5644 tests, 99.77% were successful. Failing test(s): innodb.row_size_error_log_warnings_3 disks.disks_notembedded type_test.type_test_double maria.maria-purge unit.conc_connection main.func_json_notembedded riscv64 Completed: Failed 6/6289 tests, 99.90% were successful. Failing test(s): disks.disks_notembedded type_test.type_test_double unit.conc_connection ppc64el Completed: Failed 9/6291 tests, 99.86% were successful. Failing test(s): disks.disks_notembedded type_test.type_test_double maria.maria-purge unit.conc_connection armhf Too many failed: Failed 19/1767 tests, 98.92% were successful. Failing test(s): federated.federatedx_versioning spider/bg.basic_sql spider/bg.spider_fixes_part spider.basic_sql spider.spider_fixes_part versioning.online versioning.partition period.versioning arm64 completed: Failed 9/6291 tests, 99.86% were successful. Failing test(s): disks.disks_notembedded type_test.type_test_double maria.maria-purge unit.conc_connection amd64 Completed: Failed 9/6291 tests, 99.86% were successful. Failing test(s): disks.disks_notembedded type_test.type_test_double maria.maria-purge unit.conc_connection As you can see, these seem to affect all architectures: - maria.maria-purge: MDEV-31349, easily repeatable failure on all architectures on Debian CI, Salsa CI and Launchpad - type_test.type_test_double: MDEV-22243 wasn't actually fixed yet, and failure visible on all architectures and platforms tested - unit.conc_connection: CONC-640/MDEV-30885 visible on all architectures tested - disk.disk{_notembedded}: MDEV-21587 Of these, Daniel Black is currently fixing the item 4 in the list (disk.disk) at https://github.com/MariaDB/server/pull/2777. For the items 1-3 help would be needed. The armhf situation does not map into a single MDEV. Do we have any Raspberry Pi enthusiasts here who would like to run the mariadb-test-run suite on their Pi and try to fix some of the failures? Overall it is nice to see nearly 6k tests pass on so many architectures! The Debian builds at https://buildd.debian.org/status/package.php?mariadb only run post-build the main suite (~1060 tests), and some architectures don't run any tests (e.g. m68k), just the build, but still nice to see the architecture support situation being very good at the moment for 10.11.5 across official and in-official Debian architectures as well :) Many thanks in particular to Daniel, Marko, Sergei, Tuukka, Faustin and Andrew for helping to hunt for fixes to the MariaDB code and tests to have wide architecture support! - Otto

1 0

Re: 132c56f1325: online alter: extract the source to a separate file
by Sergei Golubchik 29 Sep '23

29 Sep '23

Hi, Nikita, On Sep 29, Nikita Malyavin wrote: > revision-id: 132c56f1325 (mariadb-11.2.1-8-g132c56f1325) > parent(s): a71b13433f9 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-09-23 11:46:36 +0400 > message: > > online alter: extract the source to a separate file > > Move all the functions dedicated to online alter to a newly created > online_alter.cc. Copyright/GPL header in both new files, please. Also one comment below. > With that, make many functions static and simplify the static functions > naming. > > Also, rename binlog_log_row_online_alter -> online_alter_log_row. > > diff --git a/sql/log_event.h b/sql/log_event.h > index 848ad0cead9..35142e034f3 100644 > --- a/sql/log_event.h > +++ b/sql/log_event.h > @@ -729,19 +729,6 @@ enum Log_event_type > }; > > > -/* > - Bit flags for what has been writing to cache. Used to > - discard logs with table map events but not row events and > - nothing else important. This is stored by cache. > -*/ > - > -enum enum_logged_status > -{ > - LOGGED_TABLE_MAP= 1, > - LOGGED_ROW_EVENT= 2, > - LOGGED_NO_DATA= 4, > - LOGGED_CRITICAL= 8 > -}; > > static inline bool LOG_EVENT_IS_QUERY(enum Log_event_type type) > { > @@ -979,6 +966,19 @@ typedef struct st_print_event_info > } PRINT_EVENT_INFO; > #endif > > +/** > + Bit flags for what has been writing to cache. Used to > + discard logs with table map events but not row events and > + nothing else important. This is stored by cache. > +*/ > +enum enum_logged_status > +{ > + LOGGED_TABLE_MAP= 1, > + LOGGED_ROW_EVENT= 2, > + LOGGED_NO_DATA= 4, > + LOGGED_CRITICAL= 8 > +}; > + why did you move it? looks unintentional. Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: a71b13433f9: MDEV-32126 Assertion fails upon online ALTER and binary log enabled
by Sergei Golubchik 29 Sep '23

29 Sep '23

Hi, Nikita, On Sep 29, Nikita Malyavin wrote: > revision-id: a71b13433f9 (mariadb-11.2.1-7-ga71b13433f9) > parent(s): 883e7e5f7e7 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-09-22 17:55:14 +0400 > message: > > MDEV-32126 Assertion fails upon online ALTER and binary log enabled > > Assertion `!writer.checksum_len || writer.remains == 0' fails upon > concurrent online ALTER and transactions with failing statements and binary > log enabled. > Also another assertion, `pos != (~(my_off_t) 0)', fails in my_seek, upon > reinit_io_cache, on a simplified test. This means that IO_CACHE wasn't > properly initialized, or had an error before. > > The overall problem is a deep interference with the effect of an installed > binlog_hton. As I wrote in Jira, please, provide a better explanation here. I had to debug the issue myself and as far as I can see, the problem happens because you register binlog_hton for a transaction, but doesn't prepare the cache manager for it. This can be fixed simply with @@ -2279,6 +2279,12 @@ int binlog_log_row_online_alter(TABLE* table, const ucha> trans_register_ha(thd, false, binlog_hton, 0); if (thd->in_multi_stmt_transaction_mode()) trans_register_ha(thd, true, binlog_hton, 0); + if (mysql_bin_log.is_open()) + { + binlog_cache_mngr *cache_mngr= thd->binlog_get_cache_mngr(); + if (cache_mngr && cache_mngr->trx_cache.get_prev_position() == MY_OFF_T_UNDEF) + thd->binlog_set_stmt_begin(); + } } which solves all cases from the bug report. Don't get me wrong, I don't suggest this fix, it's kind of hackish, and I liked how other pieces of the puzzle fell into a place after a separate handlerton, like savepoints and online_alter_cache_list. I'm just saying that "magical interference with the binlog_hton" doesn't explain the problem. At all. > Solution: > Extract online alter operations into a separate handlerton. > > --- a/sql/CMakeLists.txt > +++ b/sql/CMakeLists.txt > @@ -218,6 +218,8 @@ MYSQL_ADD_PLUGIN(partition ha_partition.cc STORAGE_ENGINE DEFAULT STATIC_ONLY > RECOMPILE_FOR_EMBEDDED) > MYSQL_ADD_PLUGIN(sql_sequence ha_sequence.cc STORAGE_ENGINE MANDATORY STATIC_ONLY > RECOMPILE_FOR_EMBEDDED) > +MYSQL_ADD_PLUGIN(online_alter_log log.cc STORAGE_ENGINE MANDATORY STATIC_ONLY > +NOT_EMBEDDED) better add it next to builtin_maria_binlog_plugin, instead of via MYSQL_ADD_PLUGIN. > ADD_LIBRARY(sql STATIC ${SQL_SOURCE}) > MAYBE_DISABLE_IPO(sql) > diff --git a/sql/log.cc b/sql/log.cc > index ee78844464e..132ec254adf 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -12522,3 +12512,61 @@ void wsrep_register_binlog_handler(THD *thd, bool trx) > } > > #endif /* WITH_WSREP */ > + > +static int online_alter_close_connection(handlerton *hton, THD *thd) > +{ > + DBUG_ASSERT(thd->online_alter_cache_list.empty()); > + return 0; > +} > + > +handlerton *online_alter_hton; > + > +int online_alter_log_init(void *p) > +{ > + online_alter_hton= (handlerton *)p; > + online_alter_hton->db_type= DB_TYPE_ONLINE_ALTER; > + online_alter_hton->savepoint_offset= sizeof(my_off_t); > + online_alter_hton->close_connection= online_alter_close_connection; > + > + online_alter_hton->savepoint_set= // Done by online_alter_savepoint_set > + [](handlerton *, THD *, void *){ return 0; }; > + online_alter_hton->savepoint_rollback= // Done by online_alter_savepoint_rollback > + [](handlerton *, THD *, void *){ return 0; }; > + online_alter_hton->savepoint_rollback_can_release_mdl= > + [](handlerton *hton, THD *thd){ return true; }; > + > + online_alter_hton->commit= [](handlerton *, THD *thd, bool all) > + { return binlog_online_alter_end_trans(thd, all, true); }; > + online_alter_hton->rollback= [](handlerton *, THD *thd, bool all) > + { return binlog_online_alter_end_trans(thd, all, false); }; > + online_alter_hton->commit_by_xid= [](handlerton *hton, XID *xid) > + { return binlog_online_alter_end_trans(current_thd, true, true); }; > + online_alter_hton->rollback_by_xid= [](handlerton *hton, XID *xid) > + { return binlog_online_alter_end_trans(current_thd, true, false); }; should it be so? commit_by_xid/rollback_by_xid take XID as an argument ad you ignore it. If commit_by_xid/rollback_by_xid shouldn't be called for online alter, don't set them at all. Or set to something that always fails (and add an assert too). > + > + online_alter_hton->drop_table= [](handlerton *, const char*) { return -1; }; yeah. Like this. > + online_alter_hton->flags= HTON_NOT_USER_SELECTABLE | HTON_HIDDEN > + | HTON_NO_ROLLBACK; better remove HTON_NO_ROLLBACK. As far as I can see it doesn't do anything here anyway, so the only effect it has - everyone reading the code will stop and thing "why HTON_NO_ROLLBACK is here?" Takes at least 5-10 minutes to look through all places where HTON_NO_ROLLBACK is tested to see that online_alter_hton cannot be there. > + return 0; > +} > + > +struct st_mysql_storage_engine online_alter_storage_engine= > +{ MYSQL_HANDLERTON_INTERFACE_VERSION }; > + > +maria_declare_plugin(online_alter_log) > +{ > + MYSQL_STORAGE_ENGINE_PLUGIN, > + &online_alter_storage_engine, > + "online_alter_log", > + "MariaDB PLC", "plc" lowercase > + "This is a pseudo storage engine to represent the online alter log in a transaction", > + PLUGIN_LICENSE_GPL, > + online_alter_log_init, > + NULL, > + 0x0100, // 1.0 > + NULL, // no status vars > + NULL, // no sysvars > + "1.0", > + MariaDB_PLUGIN_MATURITY_STABLE > +} > +maria_declare_plugin_end; Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: f9c1546bd7d: MDEV-32123: require_secure_transport doesn't allow TCP connections
by Sergei Golubchik 28 Sep '23

28 Sep '23

Hi, Dmitry, On Sep 28, Dmitry Shulga wrote: > revision-id: f9c1546bd7d (mariadb-11.2.1-6-gf9c1546bd7d) > parent(s): c4cbafe0c4c > author: Dmitry Shulga > committer: Dmitry Shulga > timestamp: 2023-09-28 16:20:36 +0700 > message: > > MDEV-32123: require_secure_transport doesn't allow TCP connections > > diff --git a/mysql-test/main/require_secure_transport.test b/mysql-test/main/require_secure_transport.test > index e238e732423..4ad18afb96d 100644 > --- a/mysql-test/main/require_secure_transport.test > +++ b/mysql-test/main/require_secure_transport.test > @@ -5,11 +5,16 @@ SET GLOBAL require_secure_transport=ON; > --error ER_SECURE_TRANSPORT_REQUIRED > connect without_ssl,localhost,root,,,,,TCP NOSSL; > --enable_query_log > + > +connect with_ssl,localhost,root,,,,,TCP SSL; where's a check that a connection was established successfully and it indeed uses SSL with require_secure_transport enabled? also, you didn't add any tests for --require-secure-transport=1 > +disconnect with_ssl; > + > connection default; > SET GLOBAL require_secure_transport=OFF; > --disable_query_log > connect without_ssl,localhost,root,,,,,TCP NOSSL; > --enable_query_log > disconnect without_ssl; > + > connection default; > DROP TABLE t1; > diff --git a/sql/sql_acl.cc b/sql/sql_acl.cc > index 1f00a8670a6..3ce1c363c86 100644 > --- a/sql/sql_acl.cc > +++ b/sql/sql_acl.cc > @@ -14519,6 +14546,22 @@ bool acl_authenticate(THD *thd, uint com_change_user_pkt_len) > > if (initialized) // if not --skip-grant-tables > { > + /* > + Check whether the option require_secure_transport is on and in case > + it is true that the secured connection type is used, that is either > + unix socket or named pipe or ssl is in use. > + */ > + if(check_require_secured_transport(thd)) > + { > + Host_errors errors; > + errors.m_ssl= 1; > + inc_host_errors(mpvio.auth_info.thd->security_ctx->ip, &errors); > + status_var_increment(thd->status_var.access_denied_errors); > + my_error(ER_SECURE_TRANSPORT_REQUIRED, MYF(0)); > + > + DBUG_RETURN(1); > + } No, no. This is wrong. The point of "require_secure_transport" is to make sure that *everything* is sent over the secure transport. In particular, passwords must never be sent if the transport is not secure. You cannot check for it after the authentication, you must abort the connection is soon as possible. (technically, the server can do nothing to prevent the client from sending passwords in plain-text over the insecure connection. but practically clients use mysql_native_password by default, so the password won't leak) Regards, Sergei Chief Architect, MariaDB Server and security(a)mariadb.org

1 0

Re: [PATCH] MDEV-31872 Deprecate ENCODE()/DECODE()
by Kristian Nielsen 26 Sep '23

26 Sep '23

Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: > From: Sergei Golubchik <serg(a)mariadb.org> > @@ -2724,6 +2725,10 @@ bool Item_func_encode::seed() > hash_password(rand_nr, key->ptr(), key->length()); > sql_crypt.init(rand_nr); > > + push_warning_printf(thd, Sql_condition::WARN_LEVEL_WARN, ER_WARN_DEPRECATED_SYNTAX, > + ER_THD(thd, ER_WARN_DEPRECATED_SYNTAX_NO_REPLACEMENT), > + func_name_cstring().str); // since 11.3 > + NAK. As I already explained in the MDEV-31872, deprecating these functions is not appropriate and will hurt users, not help them. Don't do this. You need to see this from the user's point of view. She is a DBA, she has an application with columns encoded with ENCODE(). What do you expect her to do, rewrite the application because you or someone does not like these functions? That's very disrespectful. If you want to give a warning for these functions, then you need to create an SQL mode or some other handle for the DBA to disable the warning and keep their application running. And the warning should explain why it's there, and what should be done to avoid it. And you need to carefully justify how burdening the DBA with this task of tracking down the source of the new warning and the way to remove it, is worth the benefit. And deprecation suggests eventual removal, which is even worse, please don't do that. You haven't even updated the documentation https://mariadb.com/kb/en/encode/ to eg. say "encode" or "scramble" instead of encrypt. "base64_encode" also does not do encryption, nor is rot13 cryptographically safe, but that doesn't justify removing them either. Most columns in mariadb databases are not encrypted or encoded at all, how is ENCODE() any less secure that that? - Kristian.

3 10

Re: 025ca90e5df: MDEV-32101 CREATE PACKAGE [BODY] for sql_mode=DEFAULT
by Sergei Golubchik 22 Sep '23

22 Sep '23

Hi, Alexander, On Sep 15, Alexander Barkov wrote: > revision-id: 025ca90e5df (mariadb-11.0.1-231-g025ca90e5df) > parent(s): 8ad1e26b1ba > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-09-05 17:33:27 +0400 > message: > > MDEV-32101 CREATE PACKAGE [BODY] for sql_mode=DEFAULT > > This patch adds PACKAGE support with SQL/PSM dialect for sql_mode=DEFAULT: > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy > index 7452f6a083a..22c60bd4277 100644 > --- a/sql/sql_yacc.yy > +++ b/sql/sql_yacc.yy > @@ -18454,6 +18518,22 @@ drop_routine: > if (Lex->stmt_drop_procedure($3, $4)) > MYSQL_YYABORT; > } > + | DROP PACKAGE_MARIADB_SYM opt_if_exists sp_name > + { > + LEX *lex= Lex; > + lex->set_command(SQLCOM_DROP_PACKAGE, $3); > + if (unlikely(lex->sphead)) > + my_yyabort_error((ER_SP_NO_DROP_SP, MYF(0), "PACKAGE")); > + lex->spname= $4; > + } > + | DROP PACKAGE_MARIADB_SYM BODY_MARIADB_SYM opt_if_exists sp_name > + { > + LEX *lex= Lex; > + lex->set_command(SQLCOM_DROP_PACKAGE_BODY, $4); > + if (unlikely(lex->sphead)) > + my_yyabort_error((ER_SP_NO_DROP_SP, MYF(0), "PACKAGE BODY")); > + lex->spname= $5; > + } > ; at least here and in SHOW you can use package: PACKAGE_MARIADB_SYM | PACKAGE_ORACLE_SYM ; and then move drop_routine out of ifdef > > > @@ -18498,6 +18578,248 @@ create_routine: > (Item_result) $8, $10)) > MYSQL_YYABORT; > } > + > + | create_or_replace definer_opt PACKAGE_MARIADB_SYM > + opt_if_not_exists sp_name opt_create_package_chistics_init > + remember_name > + { > + sp_package *pkg; > + if (unlikely(!(pkg= Lex-> > + create_package_start(thd, > + SQLCOM_CREATE_PACKAGE, > + &sp_handler_package_spec, > + $5, $1 | $4)))) > + MYSQL_YYABORT; > + pkg->set_c_chistics(Lex->sp_chistics); > + } > + opt_sqlpsm_package_specification_element_list END > + remember_end_opt > + { > + if (unlikely(Lex->create_package_finalize(thd, $5, NULL, $7, $11))) > + MYSQL_YYABORT; > + } > + | create_or_replace definer_opt PACKAGE_MARIADB_SYM BODY_MARIADB_SYM > + opt_if_not_exists sp_name opt_create_package_chistics_init > + remember_name > + { > + sp_package *pkg; > + if (unlikely(!(pkg= Lex-> > + create_package_start(thd, > + SQLCOM_CREATE_PACKAGE_BODY, > + &sp_handler_package_body, > + $6, $1 | $5)))) > + MYSQL_YYABORT; > + pkg->set_c_chistics(Lex->sp_chistics); > + Lex->sp_block_init(thd); > + } > + sqlpsm_package_implementation_declare_section > + { > + if (unlikely(Lex->sp_block_with_exceptions_finalize_declarations(thd))) > + MYSQL_YYABORT; > + } > + sqlpsm_package_implementation_executable_section > + { > + $10.hndlrs+= $12.hndlrs; > + if (unlikely(Lex->sp_block_finalize(thd, $10))) > + MYSQL_YYABORT; > + } > + remember_end_opt > + { > + if (unlikely(Lex->create_package_finalize(thd, $6, NULL, $8, $14))) > + MYSQL_YYABORT; > + } this can also use `package` rule as above and be moved outside of %ifdef. you'll only need %ifdef MARIADB sp_tail_is: _empty ; opt_sp_name: _empty { $$= NULL; } ; %else sp_tail_is: IS | AS ; opt_sp_name: _empty { $$= NULL; } | sp_name { $$= $1; } ; %endif > + ; > + > + > +opt_sqlpsm_package_specification_element_list: > + _empty > + | sqlpsm_package_specification_element_list > + ; and so on, this can be called simply opt_package_specification_element_list if placed under %ifdef MARIADB > + > +sqlpsm_package_specification_element_list: > + sqlpsm_package_specification_element > + | sqlpsm_package_specification_element_list > + sqlpsm_package_specification_element > + ; ... Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: 14a12f98eb8: MDEV-15935 Clean up string sys_var types.
by Sergei Golubchik 14 Sep '23

14 Sep '23

Hi, Yuchen, ok to push, after making one change (const) below. On Sep 13, Yuchen Pei wrote: > revision-id: 14a12f98eb8 (mariadb-11.0.1-232-g14a12f98eb8) > parent(s): e987b9350cb > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-09-13 11:50:41 +1000 > message: > > MDEV-15935 Clean up string sys_var types. > > Merge sys_var_charptr with sys_var_charptr_base, as well as merge > Sys_var_session_lexstring into Sys_var_lexstring. Also refactored > update methods of sys_var_charptr accordingly. > > Because the class is more generic, session_update() calls > sys_var_charptr::session_update() which does not assume a buffer field > associated with THD, but instead call strdup/free, we get rid of > THD::default_master_connection_buff accordingly You could've mentioned here "this also makes THD smaller by 192 bytes, and there can be many thousands of concurrent THDs" I mean, it's a good result, one of the benefits of your cleanup. > diff --git a/sql/sys_vars.inl b/sql/sys_vars.inl > index 31ebbae01ca..5bac91a0db3 100644 > --- a/sql/sys_vars.inl > +++ b/sql/sys_vars.inl > @@ -492,14 +492,18 @@ class Sys_var_mybool: public Sys_var_typelib > Backing store: char* > > @note > - This class supports only GLOBAL variables, because THD on destruction > - does not destroy individual members of SV, there's no way to free > - allocated string variables for every thread. > + > + Note that the memory management for SESSION_VAR's is manual, the > + value must be strdup'ed in THD::init() and freed in > + plugin_thdvar_cleanup(). TODO: it should be done automatically when > + we'll have more session string variables to justify it. Maybe some > + kind of a loop over all variables, like sys_var_end() in set_var.cc? > */ > -class Sys_var_charptr_base: public sys_var > +class Sys_var_charptr: public sys_var > { > + size_t max_length= 2000; Should be const, right? No need to store max_length in every Sys_var_charptr instance, if you aren't going to modify it per instance. > public: > - Sys_var_charptr_base(const char *name_arg, > + Sys_var_charptr(const char *name_arg, > const char *comment, int flag_args, ptrdiff_t off, size_t size, > CMD_LINE getopt, > const char *def_val, PolyLock *lock=0, Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: 9db04b3dc73: MDEV-15935 Adding global/session system var redirect_url
by Sergei Golubchik 13 Sep '23

13 Sep '23

Hi, Yuchen, Thanks. Ok to push both commits into preview-11.3-preview branch On Sep 13, Yuchen Pei wrote: > revision-id: 9db04b3dc73 (mariadb-11.0.1-233-g9db04b3dc73) > parent(s): 14a12f98eb8 > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-09-13 11:50:42 +1000 > message: > > MDEV-15935 Adding global/session system var redirect_url > > Adding a global/session var `redirect_url' of string type. The initial > value is empty. Can be supplied in mysqld with --redirect-url or set > in --init-connect. A valid redirect_url should be of the format > > {mysql,mariadb}://host[:port] > > where <host> is an arbitrary string not containing colons, and <port> > is a number between 0 and 65535 inclusive. > > The variable will be used by the server to notify clients that they > should connect to another server, specified by the value of the > variable, if not empty. > > The notification is done by the inclusion of the variable in > session_track_system_variable. Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 8ac17273b14: MDEV-15935 Adding global/session system var redirect_url
by Sergei Golubchik 13 Sep '23

13 Sep '23

Hi, Yuchen, Right. This is pretty much the whole feature. Mostly ok, few minor comments below. On Sep 08, Yuchen Pei wrote: > revision-id: 8ac17273b14 (mariadb-11.0.1-104-g8ac17273b14) > parent(s): 365faf7229f > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-09-08 14:46:50 +1000 > message: > > MDEV-15935 Adding global/session system var redirect_url > > Adding a global/session var `redirect_url' of string type. The initial > value is empty. Can be supplied in mysqld with --redirect-url or set > in --init-connect. A valid redirect_url should be of the format > > {mysql,mariadb}://host[:port] > > where <host> is an arbitrary string not containing colons, and <port> > is a number between 0 and 65535 inclusive. > > The variable will be used by the server to notify clients that they > should connect to another server, specified by the value of the > variable, if not empty. > > The notification is done by the inclusion of the variable in > session_track_system_variable. > > diff --git a/mysql-test/suite/sys_vars/r/sysvars_server_embedded.result b/mysql-test/suite/sys_vars/r/sysvars_server_embedded.result > index fa3bf6b4354..85dc924fe5f 100644 > --- a/mysql-test/suite/sys_vars/r/sysvars_server_embedded.result > +++ b/mysql-test/suite/sys_vars/r/sysvars_server_embedded.result > @@ -3232,6 +3232,16 @@ NUMERIC_BLOCK_SIZE 1 > ENUM_VALUE_LIST NULL > READ_ONLY NO > COMMAND_LINE_ARGUMENT REQUIRED > +VARIABLE_NAME REDIRECT_URL > +VARIABLE_SCOPE SESSION > +VARIABLE_TYPE VARCHAR > +VARIABLE_COMMENT URL to redirect to, if not empty see if you can make it a bit more descriptive. if not, ok as is. hmm, and a second thought. it's in the sysvars_server_embedded test. it doesn't make sense there (embedded doesn't accept connections, and doesn't have session tracking, iirc), so, may be hide it under #ifndef EMBEDDED_LIBRARY ? > +NUMERIC_MIN_VALUE NULL > +NUMERIC_MAX_VALUE NULL > +NUMERIC_BLOCK_SIZE NULL > +ENUM_VALUE_LIST NULL > +READ_ONLY NO > +COMMAND_LINE_ARGUMENT REQUIRED > VARIABLE_NAME REQUIRE_SECURE_TRANSPORT > VARIABLE_SCOPE GLOBAL > VARIABLE_TYPE BOOLEAN > diff --git a/sql/sql_class.h b/sql/sql_class.h > index cf2b5b96a22..c86a76f6a08 100644 > --- a/sql/sql_class.h > +++ b/sql/sql_class.h > @@ -878,6 +878,7 @@ typedef struct system_variables > > Time_zone *time_zone; > char *session_track_system_variables; > + char *redirect_url; > > /* Some wsrep variables */ > ulonglong wsrep_trx_fragment_size; > diff --git a/sql/sql_plugin.cc b/sql/sql_plugin.cc > index 4724e2200e6..70aba9cc4d5 100644 > --- a/sql/sql_plugin.cc > +++ b/sql/sql_plugin.cc > @@ -3307,6 +3309,10 @@ void plugin_thdvar_init(THD *thd) > #ifndef EMBEDDED_LIBRARY > thd->session_tracker.sysvars.init(thd); > #endif > + thd->variables.redirect_url= > + my_strdup(PSI_INSTRUMENT_ME, global_system_variables.redirect_url, and here, key_memory_Sys_var_charptr_value > + MYF(MY_WME | MY_THREAD_SPECIFIC)); > + > DBUG_VOID_RETURN; > } > > @@ -3375,6 +3381,8 @@ void plugin_thdvar_cleanup(THD *thd) > #ifndef EMBEDDED_LIBRARY > thd->session_tracker.sysvars.deinit(thd); > #endif > + my_free(thd->variables.redirect_url); > + thd->variables.redirect_url= 0; also, perhaps in cleanup_variables() ? > my_free((char*) thd->variables.default_master_connection.str); > thd->variables.default_master_connection.str= 0; > thd->variables.default_master_connection.length= 0; > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index 064c60ab2b0..a93e68f40d7 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -1343,6 +1343,61 @@ Sys_init_connect( > DEFAULT(""), &PLock_sys_init_connect, NOT_IN_BINLOG, > ON_CHECK(check_init_string)); > > +/** > + Validate a redirect_url string. > + > + A valid string is either empty, or of the format mysql://host[:port], > + where host is an arbitrary string without any colon ':'. > + > + @param str A string to validate > + @param len Length of the string > + @retval false The string is valid > + @retval true The string is invalid > +*/ > +static bool sysvar_validate_redirect_url(sys_var *, THD *, set_var *var) > +{ > + char *str= var->save_result.string_value.str; > + size_t len= var->save_result.string_value.length; > + /* Empty string is valid */ > + if (len == 0) > + return false; > + const char* end= str + len; > + if (!strncmp(str, "mysql://", strlen("mysql://"))) use !strncmp(str, STRING_WITH_LEN("mysql://")) avoids no strlen and no need to write the string twice. same below. > + str+= strlen("mysql://"); ah. Then, I'd suggest instead LEX_CSTRING mysql_prefix={STRING_WITH_LEN("mysql://")}; if (!strncmp(str, mysql_prefix.str, mysql_prefix.length)) str+= mysql_prefix.length; > + else if (!strncmp(str, "mariadb://", strlen("mariadb://"))) > + str+= strlen("mariadb://"); > + else > + return true; > + /* Host name cannot be empty */ > + if (str == end) > + return true; > + /* Find the colon, if any */ > + while (str < end && *str != ':') > + str++; > + /* Found colon */ > + if (str < end) > + { > + /* Should have at least one number after the colon */ > + if (str + 1 == end) > + return true; > + int p= 0; > + while (str < end && isdigit(*++str)) > + p= p * 10 + (*str - '0'); > + /* Should be all numbers after the colon */ > + if (str < end) > + return true; > + /* The port number should be between 0 and 65535 */ > + if (p > 65535) > + return true; check for overflow too. you can either put the p>65535 check inside the loop or at the end check that, like, (str - colon_ptr) < 7. > + } > + return false; > +} > + > +static Sys_var_charptr Sys_redirect_url( > + "redirect_url", "URL to redirect to, if not empty", > + SESSION_VAR(redirect_url), CMD_LINE(REQUIRED_ARG), DEFAULT(""), > + NO_MUTEX_GUARD, NOT_IN_BINLOG, sysvar_validate_redirect_url, 0, 0); By convention, write ON_CHECK(sysvar_validate_redirect_url). And you don't need 0's at the end, so just NO_MUTEX_GUARD, NOT_IN_BINLOG, ON_CHECK(sysvar_validate_redirect_url)); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 4

Re: [MariaDB commits] [PATCH] MDEV-31872: Add --old_mode=COMPAT_DISCOURAGED
by Sergei Golubchik 12 Sep '23

12 Sep '23

Hi, Kristian, On Sep 12, Kristian Nielsen wrote: > > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index e25d293bbb7..18d08996aa8 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -3863,12 +3863,13 @@ static const char *old_mode_names[]= > "IGNORE_INDEX_ONLY_FOR_JOIN", // deprecated since 11.3 > "COMPAT_5_1_CHECKSUM", // deprecated since 11.3 > "LOCK_ALTER_TABLE_COPY", // deprecated since 11.3 > + "COMPAT_DISCOURAGED", > 0 > }; > > void old_mode_deprecated_warnings(THD *thd, ulonglong v) > { > - v &= ~OLD_MODE_DEFAULT_VALUE; > + v &= ~(OLD_MODE_DEFAULT_VALUE | OLD_MODE_COMPAT_DISCOURAGED); No, this is wrong. old-mode *by definition* is something that is a temporary workaround for something new, old-mode values are by design deprecated from the day one. So, OLD_MODE_COMPAT_DISCOURAGED shouldn't be on the line above, and thus it'll itself be deprecated and will cause a deprecation warning. Which is rather silly, as I wrote in a previous email, and that's why an old-mode is not a good solution for turning deprecated warnings away. > for (uint i=0; old_mode_names[i]; i++) > if ((1ULL<<i) & v) > { > diff --git a/sql/item_strfunc.cc b/sql/item_strfunc.cc > index 4b96271c331..339430a2794 100644 > --- a/sql/item_strfunc.cc > +++ b/sql/item_strfunc.cc > @@ -2720,7 +2720,16 @@ bool Item_func_encode::seed() > hash_password(rand_nr, key->ptr(), key->length()); > sql_crypt.init(rand_nr); > > - warn_deprecated<1103>(current_thd, func_name_cstring().str); > + if (!(current_thd->variables.old_behavior & OLD_MODE_COMPAT_DISCOURAGED)) > + { > + /* > + This function should not be removed despite the deprecation warning. > + That would cause problems for users who can then no longer access > + columns in their database that might have been inserted as > + ENCODE(str,pass_str). > + */ > + warn_deprecated<1103>(current_thd, func_name_cstring().str); > + } No, this is exactly what I wanted to prevent by adding a warn_deprecated<>() wrapper. If something is giving a warning "... is deprecated and will be removed" - it *has to be removed*. We don't want to lie to users to simply scare them. If we promise to do something, we should do it. If we aren't going to remove ENCODE/DECODE functions, we shouldn't deprecate them at all. Which is the result I'm leaning to right now. > > return FALSE; > } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 75ea482c4fc: MDEV-30973 GTID in deadlock output
by Sergei Golubchik 11 Sep '23

11 Sep '23

Hi, Aleksey, On Sep 11, Aleksey Midenkov wrote: > revision-id: 75ea482c4fc (mariadb-11.0.1-230-g75ea482c4fc) > parent(s): fe62f70fb4b > author: Aleksey Midenkov > committer: Aleksey Midenkov > timestamp: 2023-09-01 20:30:24 +0300 > message: > > MDEV-30973 GTID in deadlock output > > Example: > > TRANSACTION 40, ACTIVE 0 sec starting index read > mysql tables in use 1, locked 1 > LOCK WAIT 2 lock struct(s), heap size 1152, 1 row lock(s) > MySQL thread id 23, OS thread handle 140201491547840, query id 310, GTID 3-1-1 Updating > delete from tu where a = 9999 and b = 'xxxx' > 2023-04-12 19:53:58 22 [Note] InnoDB: *** (1) WAITING FOR THIS LOCK TO BE GRANTED: > > Deterministic slave_deadlock.test > > Instead of "MDEV-10962 Deadlock with 3 concurrent DELETEs by unique key" > this one uses "MDEV-22698 Deadlock on concurrent acquisition from multiple indexes" > > --- > mysql-test/suite/rpl/t/slave_deadlock.test | 93 ++++++++++++++++++++++++++++++ > sql/sql_show.cc | 10 ++++ please move slave_deadlock.result from the following commit to this one. otherwise ok Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Implementing Binlog GTID Indexes (MDEV-4991)
by Kristian Nielsen 10 Sep '23

10 Sep '23

I am implementing Binlog GTID Indexes, to fix an old performance regression for GTID when slaves connect to the master. I now have a solid design and a working prototype, so I wanted to describe the work to encourage early comments and suggestions. I'm hoping this will make it to the next release (where is the place to actively follow/participate in the release planning?) Ever since I implemented GTID, connecting slaves needed to sequentially scan the leading portion of one binlog file to locate the GTID position to start at. This can be slow as binlog files default to 1GB in size, even worse so when binlog files can now be encrypted. I files MDEV-4991 for this 10 years ago, and it is embarrassing that this has somehow remained unfixed for so long :-/ But now I'm fixing it. The current implementation is in the work-in-progress branch knielsen_mdev4991: https://github.com/MariaDB/server/tree/knielsen_mdev4991 The code is already functional, with the testsuite passing. Still missing is binlog purge, index crash recovery, async path, and general testing and cleanup. The basic idea is for each binlog file master-bin.000001 to write an index file master-bin.000001.idx. The index contains a B+-Tree in which the keys are pairs of (GTID state, binlog offset). A connecting slave's GTID position can be looked up quickly in the tree to find the corresponding binlog offset (and gtid_binlog_state) to start replicating at. Similarly, a non-GTID connecting slave can quickly look up its starting offset (BINLOG_GTID_POS()) to obtain the GTID position corresponding to its starting offset. The index is written out to disk concurrently with the writing of the binlog file. Since the tree is written in sorted order and append-only, the implementation is significantly simplified over a general B+-Tree, and can also be described as a Log-structed Merge Tree with a Btree search structure on top. All GTID state keys (except the first) in the index are delta-compressed, storing only the GTIDs that changed since the last record (which typically will be only one). Additionally the index is sparse, storing only say every 1-in-10 GTIDs in the index (this will be configurable of course). Using a sparse index, the disk space is reduced at the cost only of scanning a few extra events of the binlog file to find the exact position requested. Actual benchmarking of space usage is TBO, but we can expect a typical binlog file of 1 GB containing say 3 million GTIDs to require something like 10 MB of index disk space (1% increase). The file format is page-based (unlike the binlog file), so that it can be written efficiently and easily read using random-access. The index is written in parallel with the binlog file, but no extra fsync()s are needed (except one at the end of the file when the index is closed and synced to disk). The writing of the index will happen asynchroneously from the binlog background thread; this way it will have minimal impact on the performance and scalability of the binlog and the contested LOCK_log. Connecting slaves can read from the currently-being-written ("hot") GTID index by accessing internal memory buffers of pages not yet written to disk. If the server crashes, the existing binlog recovery and binlog checkpoint mechanism will be used to re-create any incomplete indexes as part of the normal recovery scan of binlog files. If somehow a binlog index should be found corrupt or missing (could be eg. old server upgrade or backup script that omits the new .idx files), the code will gracefully fall back to the old (slower) way of sequentially scanning the binlog file to locate the slave starting position. I think this will be a great (if rather late) improvement to one remaining corner of the GTID implementation that currently has sub-optimal performance, and I'm exited to get this completed and reviewed/tested and added to hopefully the next MariaDB release. - Kristian.

1 0

Re: 365faf7229f: MDEV-15935 Clean up string sys_var types.
by Sergei Golubchik 08 Sep '23

08 Sep '23

Hi, Yuchen, On Sep 08, Yuchen Pei wrote: > revision-id: 365faf7229f (mariadb-11.0.1-103-g365faf7229f) > parent(s): 4f39ae75606 > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-09-08 13:12:22 +1000 > message: > > MDEV-15935 Clean up string sys_var types. > > Merge sys_var_charptr with sys_var_charptr_base, as well as merge > Sys_var_session_lexstring into Sys_var_lexstring. Also refactored > update methods of sys_var_charptr accordingly. > > Because the class is more generic, session_update() calls > sys_var_charptr::session_update() which does not assume a buffer field > associated with THD, but instead call strdup/free, we get rid of > THD::default_master_connection_buff accordingly > diff --git a/sql/sql_class.cc b/sql/sql_class.cc > index 18e796655bd..52fa6c27112 100644 > --- a/sql/sql_class.cc > +++ b/sql/sql_class.cc > @@ -1233,10 +1233,12 @@ void THD::init() > avoid temporary tables replication failure. > */ > variables.pseudo_thread_id= thread_id; > - variables.default_master_connection.str= default_master_connection_buff; > - ::strmake(default_master_connection_buff, > - global_system_variables.default_master_connection.str, > - variables.default_master_connection.length); > + variables.default_master_connection.str= > + my_strndup(PSI_INSTRUMENT_ME, may be key_memory_Sys_var_charptr_value instead of PSI_INSTRUMENT_ME? > + global_system_variables.default_master_connection.str, > + global_system_variables.default_master_connection.length, > + MYF(MY_WME | MY_THREAD_SPECIFIC)); > + > mysql_mutex_unlock(&LOCK_global_system_variables); > > user_time.val= start_time= start_time_sec_part= 0; > diff --git a/sql/sql_plugin.cc b/sql/sql_plugin.cc > index d39589ed1ab..4724e2200e6 100644 > --- a/sql/sql_plugin.cc > +++ b/sql/sql_plugin.cc > @@ -3279,6 +3279,9 @@ void plugin_thdvar_init(THD *thd) > #ifndef EMBEDDED_LIBRARY > thd->session_tracker.sysvars.deinit(thd); > #endif > + my_free((char*) thd->variables.default_master_connection.str); > + thd->variables.default_master_connection.str= 0; > + thd->variables.default_master_connection.length= 0; shouldn't this go inside cleanup_variables() ? > > thd->variables= global_system_variables; > > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index 4843450352e..064c60ab2b0 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -1355,12 +1355,11 @@ static bool check_master_connection(sys_var *self, THD *thd, set_var *var) > return false; > } > > -static Sys_var_session_lexstring Sys_default_master_connection( > +static Sys_var_lexstring Sys_default_master_connection( > "default_master_connection", > "Master connection to use for all slave variables and slave commands", > - SESSION_ONLY(default_master_connection), > - NO_CMD_LINE, > - DEFAULT(""), MAX_CONNECTION_NAME, ON_CHECK(check_master_connection)); > + SESSION_ONLY(default_master_connection), NO_CMD_LINE, DEFAULT(""), > + NO_MUTEX_GUARD, NOT_IN_BINLOG, ON_CHECK(check_master_connection)); some length limit would still be nice, I'd say. Doesn't have to be per variable, may be all Sys_var_lexstring could have some common limit, like, arbitrarily, 1K or 10K. > #endif > > static Sys_var_charptr_fscs Sys_init_file( > diff --git a/sql/sys_vars.inl b/sql/sys_vars.inl > index 31ebbae01ca..0f2a07d92fc 100644 > --- a/sql/sys_vars.inl > +++ b/sql/sys_vars.inl > @@ -576,6 +571,21 @@ class Sys_var_charptr_base: public sys_var > new_val= 0; > return new_val; > } > + char *session_update_prepare(set_var *var) do you need this session_update_prepare method? global_update_prepare() and global_update_finish() are used by Sys_var_sesvartrack. Getting rid of them is a task for another day :) > + { > + return update_prepare(var, MYF(MY_WME | MY_THREAD_SPECIFIC)); > + } > + bool session_update(THD *thd, set_var *var) override > + { > + char *new_val= session_update_prepare(var); > + my_free(session_var(thd, char*)); > + session_var(thd, char*)= new_val; > + return (new_val == 0 && var->save_result.string_value.str != 0); > + } > + char *global_update_prepare(set_var *var) > + { > + return update_prepare(var, MYF(MY_WME)); > + } > void global_update_finish(char *new_val) > { > if (flags & ALLOCATED) > @@ -583,50 +593,28 @@ class Sys_var_charptr_base: public sys_var > flags|= ALLOCATED; > global_var(char*)= new_val; > } > - bool global_update(THD *thd, set_var *var) > + bool global_update(THD *thd, set_var *var) override > { > - char *new_val= global_update_prepare(thd, var); > + char *new_val= global_update_prepare(var); > global_update_finish(new_val); > return (new_val == 0 && var->save_result.string_value.str != 0); > } > - void session_save_default(THD *thd, set_var *var)= 0; > - void global_save_default(THD *thd, set_var *var) > + void save_default(set_var *var) > { > char *ptr= (char*)(intptr)option.def_value; > var->save_result.string_value.str= ptr; > var->save_result.string_value.length= ptr ? strlen(ptr) : 0; > } > -}; > - > -class Sys_var_charptr: public Sys_var_charptr_base > -{ > -public: > - Sys_var_charptr(const char *name_arg, > - const char *comment, int flag_args, ptrdiff_t off, size_t size, > - CMD_LINE getopt, > - const char *def_val, PolyLock *lock=0, > - enum binlog_status_enum binlog_status_arg=VARIABLE_NOT_IN_BINLOG, > - on_check_function on_check_func=0, > - on_update_function on_update_func=0, > - const char *substitute=0) : > - Sys_var_charptr_base(name_arg, comment, flag_args, off, size, getopt, > - def_val, lock, binlog_status_arg, > - on_check_func, on_update_func, substitute) > + void session_save_default(THD *, set_var *var) override > { > - SYSVAR_ASSERT(scope() == GLOBAL); > - SYSVAR_ASSERT(size == sizeof(char *)); > + return save_default(var); no, session's default value is global's value. See any other Sys_var_xxx class, e.g. for Sys_var_mybool: void session_save_default(THD *thd, set_var *var) { var->save_result.ulonglong_value= (ulonglong)*(my_bool *)global_value_ptr(thd, 0); } You cannot have a test for that now, but let's have one in the redirect_url commit. > } > - > - bool session_update(THD *thd, set_var *var) > + void global_save_default(THD *, set_var *var) override > { > - DBUG_ASSERT(FALSE); > - return true; > + return save_default(var); > } > - void session_save_default(THD *thd, set_var *var) > - { DBUG_ASSERT(FALSE); } > }; > > - > class Sys_var_charptr_fscs: public Sys_var_charptr > { > using Sys_var_charptr::Sys_var_charptr; > diff --git a/mysql-test/suite/sys_vars/r/mdev_15935.result b/mysql-test/suite/sys_vars/r/mdev_15935.result > new file mode 100644 > index 00000000000..d20c2ad2f19 > --- /dev/null > +++ b/mysql-test/suite/sys_vars/r/mdev_15935.result > @@ -0,0 +1,12 @@ > +# > +# test that the global/session-only enforcement is still in place despite the merge of string sys_var classes > +# > +set session init_connect="whatever"; > +ERROR HY000: Variable 'init_connect' is a GLOBAL variable and should be set with SET GLOBAL > +set session wsrep_auto_increment_control="whatever"; > +ERROR HY000: Variable 'wsrep_auto_increment_control' is a GLOBAL variable and should be set with SET GLOBAL > +set global default_master_connection="whatever"; > +ERROR HY000: Variable 'default_master_connection' is a SESSION variable and can't be used with SET GLOBAL not sure you need that, it's checked in the very base class, you could've have affected that I'd say it's enough if existing tests for Sys_var_session_lexstring and Sys_var_sesvartrack variables succeed and ASAN/MSAN are happy > +# > +# end of test mdev_15935 > +# Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 04b99a30a3f: MDEV-15935 Adding global/session var redirect_url
by Sergei Golubchik 08 Sep '23

08 Sep '23

Hi, Yuchen, On Sep 01, Yuchen Pei wrote: > revision-id: 04b99a30a3f (mariadb-11.0.1-103-g04b99a30a3f) > parent(s): 4f39ae75606 > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-08-16 11:31:16 +1000 > message: > > MDEV-15935 Adding global/session var redirect_url > > Adding a global/session var `redirect_url' of string type. The initial > value is empty. Can be supplied in mysqld with --redirect-url. A valid > redirect_url should be of the format > > {mysql,mariadb}://host[:port] > > where <host> is an arbitrary string not containing colons, and <port> > is a number between 0 and 65535 inclusive. > > diff --git a/mysql-test/main/redirect.opt b/mysql-test/main/redirect.opt move this to the sys_vars suite, please > --- /dev/null > +++ b/mysql-test/main/redirect.opt > @@ -0,0 +1 @@ > +--redirect_url=mysql://foobar ok, as long as this isn't a complete feature, but just one piece of it. but when client support will be implemented (CONC-599) and the server will actually send this to the connecting clients (MDEV-31609), this test will need to be fixed, as foobar is not something a client can be redirected to. Also, it'd be more interesting to try something like --init-connect='set @@redirect_url="mysql://foobar"' instead of the simple --redirect_url=mysql://foobar because init-connect is much more flexible, it can generate the url on the fly, it can send different users to different locations. So it'd make sense to test init-connect approach. > diff --git a/mysql-test/main/redirect.test b/mysql-test/main/redirect.test > new file mode 100644 > --- /dev/null > +++ b/mysql-test/main/redirect.test > @@ -0,0 +1,56 @@ > +--echo # > +--echo # MDEV-15935 Connection Redirection Mechanism in MariaDB Client/Server Protocol > +--echo # > + > +set @old_redirect_url=@@redirect_url; > +select @@global.redirect_url; > +set global redirect_url=default; > +select @@global.redirect_url; > +--error ER_WRONG_VALUE_FOR_VAR > +set global redirect_url="mariadb.org"; > +--error ER_WRONG_VALUE_FOR_VAR > +set global redirect_url="https://mariadb.org"; before doing all this fun, please remove redirect_url from the session_track_system_variables, or just set it to the empty string. > diff --git a/sql/sql_class.h b/sql/sql_class.h > --- a/sql/sql_class.h > +++ b/sql/sql_class.h > @@ -3585,6 +3587,7 @@ class THD: public THD_count, /* this must be first */ > */ > LEX_CSTRING connection_name; > char default_master_connection_buff[MAX_CONNECTION_NAME+1]; > + char redirect_url_buff[MAX_CONNECTION_NAME+1]; This isn't such a good idea (I'm afraid, I missed it when default_master_connection_buff was added). A couple of years ago Monty have spent significant efforts trying to make THD smaller as it has grown to be too large. Now THD keeps growing again. Let's try not to encourage it, if we can help it. > uint8 password; /* 0, 1 or 2 */ > uint8 failed_com_change_user; > bool slave_thread; > diff --git a/sql/sys_vars.inl b/sql/sys_vars.inl > --- a/sql/sys_vars.inl > +++ b/sql/sys_vars.inl > @@ -637,6 +637,129 @@ class Sys_var_charptr_fscs: public Sys_var_charptr ... > +class Sys_var_redirect_url: public Sys_var_charptr_base please, no. new classes are needed only for variables with very special, unlike anything else, behavior. redirect_url is just a session string variable. See the comment at the top of sys_vars.cc But (!) there are no other session string variables yet. So I'd recomment to fix Sys_var_charptr to support session variables. And then create redirect_url out of it. Practically it might need no changes at all, if you put explicit strmake (or strdup/free) into THD::init like you did anyway. After that, I suspect, you might be able to remove Sys_var_charptr_base, Sys_var_sesvartrack, and Sys_var_session_lexstring, because Sys_var_charptr/Sys_var_lexstring will be able to serve all those use cases. > +{ > + static const size_t max_length = MAX_CONNECTION_NAME; > +public: > + Sys_var_redirect_url(const char *name_arg, > + const char *comment, > + CMD_LINE getopt, > + const char *def_val, PolyLock *lock= 0) : > + Sys_var_charptr_base(name_arg, comment, > + SESSION_VAR(redirect_url), > + getopt, def_val, lock, > + VARIABLE_NOT_IN_BINLOG, 0, 0, 0) > + { > + global_var(LEX_CSTRING).length= strlen(def_val); > + option.var_type|= GET_STR; > + *const_cast<SHOW_TYPE*>(&show_val_type)= SHOW_LEX_STRING; > + } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: 44e692d0dbc: MDEV-29167 new db-level SHOW CREATE ROUTINE privilege
by Sergei Golubchik 06 Sep '23

06 Sep '23

Hi, Oleksandr, On Sep 04, Oleksandr Byelkin wrote: > revision-id: 44e692d0dbc (mariadb-11.0.1-228-g44e692d0dbc) > parent(s): 9cb75f333fa > author: Oleksandr Byelkin > committer: Oleksandr Byelkin > timestamp: 2023-09-04 14:12:29 +0200 > message: > > MDEV-29167 new db-level SHOW CREATE ROUTINE privilege > > -- Create general_log if CSV is enabled. > diff --git a/scripts/mariadb_system_tables_fix.sql b/scripts/mariadb_system_tables_fix.sql > --- a/scripts/mariadb_system_tables_fix.sql > +++ b/scripts/mariadb_system_tables_fix.sql > @@ -706,6 +706,9 @@ ALTER TABLE db modify Delete_history_priv enum('N','Y') COLLATE utf8mb3_general_ > > UPDATE user SET Delete_history_priv = Super_priv WHERE @had_user_delete_history_priv = 0; > > +ALTER TABLE db ADD Show_create_routine_priv enum('N','Y') COLLATE utf8mb3_general_ci NOT NULL DEFAULT 'N' AFTER Delete_history_priv; > +ALTER TABLE db modify Show_create_routine_priv enum('N','Y') COLLATE utf8mb3_general_ci NOT NULL DEFAULT 'N'; Why two statements? > + > ALTER TABLE user ADD plugin char(64) CHARACTER SET latin1 DEFAULT '' NOT NULL AFTER max_user_connections, > ADD authentication_string TEXT NOT NULL AFTER plugin; > ALTER TABLE user CHANGE auth_string authentication_string TEXT NOT NULL; > diff --git a/sql/sql_acl.cc b/sql/sql_acl.cc > --- a/sql/sql_acl.cc > +++ b/sql/sql_acl.cc > @@ -9003,7 +9008,7 @@ bool check_routine_level_acl(THD *thd, const char *db, const char *name, > NULL, db, > sctx->priv_role, > name, sph, 0))) > - no_routine_acl= !(grant_proc->privs & SHOW_PROC_ACLS); > + no_routine_acl= !(grant_proc->privs & SHOW_PROC_NO_DEF_ACLS); Oh, now I get it. All the time up until this line I thought that SHOW_PROC_NO_DEF_ACLS means "privileges that *not a definer* of the routine needs to have" And now I see that it means to "show procedure without definition". Please, rename to something, like, SHOW_PROC_EXISTANCE or SHOW_PROC_NO_BODY or even SHOW_PROC_WITHOUT_DEFINITION. But SHOW_PROC_NO_DEF is rather misleading. > } > mysql_rwlock_unlock(&LOCK_grant); > return no_routine_acl; > @@ -9237,7 +9242,7 @@ static const char *command_array[]= > "CREATE USER", "EVENT", "TRIGGER", "CREATE TABLESPACE", "DELETE HISTORY", > "SET USER", "FEDERATED ADMIN", "CONNECTION ADMIN", "READ_ONLY ADMIN", > "REPLICATION SLAVE ADMIN", "REPLICATION MASTER ADMIN", "BINLOG ADMIN", > - "BINLOG REPLAY", "SLAVE MONITOR" > + "BINLOG REPLAY", "SLAVE MONITOR", "SHOW CREATE ROUTINE" > }; > > static uint command_lengths[]= > @@ -9250,7 +9255,7 @@ static uint command_lengths[]= > 11, 5, 7, 17, 14, > 8, 15, 16, 15, > 23, 24, 12, > - 13, 13 > + 13, 13, 19 May be you'll change these two arrays into one of LEX_CSTRING's ? in a separate commit, of course. > }; > > diff --git a/sql/sql_show.cc b/sql/sql_show.cc > --- a/sql/sql_show.cc > +++ b/sql/sql_show.cc > @@ -6676,7 +6677,8 @@ int store_schema_params(THD *thd, TABLE *table, TABLE *proc_table, > DBUG_RETURN(0); > > if (!full_access) > - full_access= !strcmp(sp_user, definer.str); > + full_access= !check_db_routine_access(thd, SHOW_CREATE_ROUTINE_ACL, > + db.str, name.str, sph, TRUE); Wait, I thought we agreed that to see a routine body one has to be either the definer or granted SHOW CREATE ROUTINE. As far as I can see check_db_routine_access doesn't check for the user being the definer. > if (!full_access && > check_some_routine_access(thd, db.str, name.str, sph)) > DBUG_RETURN(0); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

3 3

Re: 82a49812507: MDEV-22597 Add views for periods in information_schema
by Sergei Golubchik 05 Sep '23

05 Sep '23

Hi, Nikita, On Aug 31, Nikita Malyavin wrote: > revision-id: 82a49812507 (mariadb-11.0.1-228-g82a49812507) > parent(s): 23b15d24eb3 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-08-30 22:49:36 +0400 > message: > > MDEV-22597 Add views for periods in information_schema > > Two new information_schema views are added: > * PERIOD table -- columns TABLE_CATALOG, TABLE_SCHEMA, TABLE_NAME, > PERIOD_NAME, START_COLUMN_NAME, END_COLUMN_NAME. > * KEY_PERIOD_USAGE -- works similar to KEY_COLUMN_USAGE, but for periods. > Columns CONSTRAINT_CATALOG, CONSTRAINT_SCHEMA, CONSTRAINT_NAME, > TABLE_CATALOG, TABLE_SCHEMA, TABLE_NAME, PERIOD_NAME > > Two new columns are added to the COLUMNS view: > IS_SYSTEM_TIME_PERIOD_START, IS_SYSTEM_TIME_PERIOD_END - contain YES/NO. > > diff --git a/mysql-test/main/information_schema-big.result b/mysql-test/main/information_schema-big.result > index 94c8d274fad..ad0451293c9 100644 > --- a/mysql-test/main/information_schema-big.result > +++ b/mysql-test/main/information_schema-big.result good, you didn't forget to update big and embedded tests > diff --git a/mysql-test/suite/period/r/i_s_notembedded.result b/mysql-test/suite/period/r/i_s_notembedded.result > new file mode 100644 > index 00000000000..879376cc8b4 > --- /dev/null > +++ b/mysql-test/suite/period/r/i_s_notembedded.result > @@ -0,0 +1,52 @@ > +select * from information_schema.periods; > +TABLE_CATALOG TABLE_SCHEMA TABLE_NAME PERIOD START_COLUMN_NAME END_COLUMN_NAME > +create or replace table t1 (id int primary key, s timestamp(6), e timestamp(6), > +period for mytime(s,e)); > +create or replace table t2 (id int primary key, s timestamp(6), e timestamp(6), > +period for mytime(s,e)) with system versioning; > +show columns from t1; > +Field Type Null Key Default Extra > +id int(11) NO PRI NULL > +s timestamp(6) NO NULL > +e timestamp(6) NO NULL > +select * from information_schema.periods where table_schema = 'test'; > +TABLE_CATALOG TABLE_SCHEMA TABLE_NAME PERIOD START_COLUMN_NAME END_COLUMN_NAME > +def test t2 SYSTEM_TIME row_start row_end I don't think so, they're INVISIBLE_SYSTEM. for all practical purposes they don't exist as columns in the table. Shouldn't be shown either. But don't remove this test, let's keep it to show that row_start/row_end are *not* shown here. And add table t3 with full standard definition of period for system_time. it should be shown all right. > +def test t2 mytime s e > +def test t1 mytime s e > +create user periods_hidden@localhost; ... > diff --git a/sql/sql_show.cc b/sql/sql_show.cc > index 1dc60a01f38..876d6c6f244 100644 > --- a/sql/sql_show.cc > +++ b/sql/sql_show.cc > @@ -6165,11 +6166,83 @@ static void store_variable_type(THD *thd, const sp_variable *spvar, > } > } > > +static int store_schema_period_record(THD *thd, TABLE_LIST *tl, > + TABLE *schema_table, > + const LEX_CSTRING *db_name, > + const LEX_CSTRING *table_name, > + const TABLE_SHARE::period_info_t &period) > +{ > + TABLE_SHARE *s= tl->table->s; > + const CHARSET_INFO *cs= system_charset_info; > + > + int err= schema_table->field[0]->store(STRING_WITH_LEN("def"), cs); > + if(!err) err= schema_table->field[1]->store(db_name, cs); > + if(!err) err= schema_table->field[2]->store(table_name, cs); > + if(!err) err= schema_table->field[3]->store(period.name, cs); well, ok. but as you've seen elsewhere these go without checks, as these calls generally cannot fail, they don't even allocate any memory. > + > + if (err) return err; > + > + int period_field= 4; > + for (auto *field: {period.start_field(s), period.end_field(s)}) > + { > +#ifndef NO_EMBEDDED_ACCESS_CHECKS > + bool col_granted= get_column_grant(thd, &tl->grant, > + db_name->str, table_name->str, > + field->field_name.str) & COL_ACLS; > + if (col_granted) > +#endif > + { > + schema_table->field[period_field]->set_notnull(); > + err= schema_table->field[period_field]->store(field->field_name, cs); > + } > + > + if (err) > + return err; > + period_field++; > + } > + > + return schema_table_store_record(thd, schema_table); > +} > + > + > +static int > +get_schema_period_records(THD *thd, TABLE_LIST *tl, > + TABLE *schema_table, //!< @ref > + //!< Show::periods_fields_info > + bool res, > + const LEX_CSTRING *db_name, > + const LEX_CSTRING *table_name) > +{ > + TABLE *table= tl->table; > + > + if (!table || (!table->s->period.name && !table->versioned())) > + return 0; > + > +#ifndef NO_EMBEDDED_ACCESS_CHECKS > + check_access(thd, SELECT_ACL, db_name->str, > + &tl->grant.privilege, 0, 0, MY_TEST(tl->schema_table)); > + if (is_temporary_table(tl)) > + { > + tl->grant.privilege|= TMP_TABLE_ACLS; > + } > +#endif > + int err= 0; > + if (table->versioned()) > + err= store_schema_period_record(thd, tl, schema_table, db_name, table_name, > + table->s->vers); > + if (!err && table->s->period.name) > + err= store_schema_period_record(thd, tl, schema_table, db_name, table_name, > + table->s->period); > + > + return err; > +} > > -static int get_schema_column_record(THD *thd, TABLE_LIST *tables, > - TABLE *table, bool res, > - const LEX_CSTRING *db_name, > - const LEX_CSTRING *table_name) > +static > +int get_schema_column_record(THD *thd, TABLE_LIST *tables, > + TABLE *table, //!< @ref Show::columns_fields_info > + bool res, > + const LEX_CSTRING *db_name, > + const LEX_CSTRING *table_name) > { > LEX *lex= thd->lex; > const char *wild= lex->wild ? lex->wild->ptr() : NullS; > @@ -7513,30 +7592,41 @@ static int get_schema_triggers_record(THD *thd, TABLE_LIST *tables, > DBUG_RETURN(0); > } > > +static int > +store_key_column_usage(TABLE *table, const LEX_CSTRING &db_name, > + const LEX_CSTRING &table_name, > + const LEX_CSTRING &key_name, > + const LEX_CSTRING &col_name) > +{ > + CHARSET_INFO *cs= system_charset_info; > + static const LEX_CSTRING def{STRING_WITH_LEN("def")}; > + const LEX_CSTRING *values[] { &def, &db_name, &key_name, > + &def, &db_name, &table_name, &col_name }; > + int err = 0; > + for (uint i = 0; i < array_elements(values) && !err; i++) > + table->field[i]->store(values[i], cs); you forgot `err=` here :) but as I wrote above, it's not needed here anyway > + return 0; > +} > > static void > store_key_column_usage(TABLE *table, const LEX_CSTRING *db_name, > const LEX_CSTRING *table_name, const char *key_name, > - size_t key_len, const char *con_type, size_t con_len, > + size_t key_len, const char *col_name, size_t col_len, > longlong idx) > { > - CHARSET_INFO *cs= system_charset_info; > - table->field[0]->store(STRING_WITH_LEN("def"), cs); > - table->field[1]->store(db_name->str, db_name->length, cs); > - table->field[2]->store(key_name, key_len, cs); > - table->field[3]->store(STRING_WITH_LEN("def"), cs); > - table->field[4]->store(db_name->str, db_name->length, cs); > - table->field[5]->store(table_name->str, table_name->length, cs); > - table->field[6]->store(con_type, con_len, cs); > + store_key_column_usage(table, *db_name, *table_name, {key_name, key_len}, > + {col_name, col_len}); pretty, but better, please, put it back. You can use the old store_key_column_usage from your get_schema_key_period_usage_record() just the same. There's no need to have two store_key_column_usage() functions. > table->field[7]->store((longlong) idx, TRUE); > } > > > -static int get_schema_key_column_usage_record(THD *thd, > - TABLE_LIST *tables, > - TABLE *table, bool res, > - const LEX_CSTRING *db_name, > - const LEX_CSTRING *table_name) > +static int > +get_schema_key_column_usage_record(THD *thd, TABLE_LIST *tables, > + TABLE *table, > + //!< @ref Show::key_column_usage_fields_info > + bool res, > + const LEX_CSTRING *db_name, > + const LEX_CSTRING *table_name) > { > DBUG_ENTER("get_schema_key_column_usage_record"); > if (res) > @@ -9703,6 +9824,18 @@ ST_FIELD_INFO key_column_usage_fields_info[]= > CEnd() > }; > > +ST_FIELD_INFO key_period_usage_fields_info[]= > +{ > + Column("CONSTRAINT_CATALOG", Catalog(), NOT_NULL, OPEN_FULL_TABLE), > + Column("CONSTRAINT_SCHEMA", Name(), NOT_NULL, OPEN_FULL_TABLE), > + Column("CONSTRAINT_NAME", Name(), NOT_NULL, OPEN_FULL_TABLE), > + Column("TABLE_CATALOG", Catalog(), NOT_NULL, OPEN_FULL_TABLE), > + Column("TABLE_SCHEMA", Name(), NOT_NULL, OPEN_FULL_TABLE), > + Column("TABLE_NAME", Name(), NOT_NULL, OPEN_FULL_TABLE), > + Column("PERIOD_NAME", Name(), NOT_NULL, OPEN_FULL_TABLE), Cannot they all be OPEN_FRM_ONLY? > + CEnd() > +}; > + > > ST_FIELD_INFO table_names_fields_info[]= > { > @@ -10208,6 +10358,8 @@ ST_SCHEMA_TABLE schema_tables[]= > {0, 0, 0, 0, 0, 0, 0, 0, 0, 0} > }; > > +static_assert(array_elements(schema_tables) == SCH_ENUM_SIZE + 1, > + "Update enum_schema_tables as well."); +1 > > int initialize_schema_table(st_plugin_int *plugin) > { Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 6

Re: 8114fb63bbb: MDEV-31684: Add timezone information to DATE_FORMAT
by Sergei Golubchik 04 Sep '23

04 Sep '23

Hi, Rucha, Few comments, after I saw the commit in the branch: On Sep 04, Rucha Deodhar wrote: > revision-id: 8114fb63bbb (mariadb-11.0.1-229-g8114fb63bbb) > parent(s): 7ba9c7fb84b > author: Rucha Deodhar > committer: Rucha Deodhar > timestamp: 2023-09-01 11:01:09 +0530 > message: > > MDEV-31684: Add timezone information to DATE_FORMAT > > Before starting to go over the format string, prepare the current time > zone information incase '%z' or '%Z' is encountered. > This information can be obtained as given below: > > A) If timezone is not set ( meaning we are working with system timezone): > Get the MYSQL_TIME representation for current time and GMT time using > current thread variable for timezone and timezone variable for UTC > respectively. This MYSQL_TIME variable will be used to calculate time > difference. Also convert current time in second to tm structure to > get system timezone information. > > B) If timezone is set as offset: > Get timezone information using current timezone information and store > in appropriate variable. > > C) If timezone is set as some place (example: Europe/Berlin) > Get timezone information by searching the timezone. During internal > timezone search, information like timeoffset from UTC and abbrevation > is stored in another relevant structure. Hence use the same information. > > diff --git a/mysql-test/main/date_formats.result b/mysql-test/main/date_formats.result > index 463cce39520..23aeffedef7 100644 > --- a/mysql-test/main/date_formats.result > +++ b/mysql-test/main/date_formats.result > @@ -573,3 +573,67 @@ time_format('01 02:02:02', '%T') > select time_format('2001-01-01 02:02:02', '%T'); > time_format('2001-01-01 02:02:02', '%T') > 02:02:02 > +# > +# Beginning of 11.3 test > +# > +# MDEV-31684: Add timezone information to DATE_FORMAT > +# > +SET @old_timezone= @@time_zone; > +SET TIME_ZONE='Japan'; > +SELECT DATE_FORMAT('2009-10-04 22:23:00', '%z %Z') AS current_timezone; > +current_timezone > ++0900 CJT > +SELECT DATE_FORMAT(CURDATE(), '%z %Z') AS current_tz; 1) please, don't use CURDATE/CURRENT_TIMESTAMP/etc. The time zone abbreviation depends on the date: $ TZ=Europe/Berlin date -d '2020-01-01 1:2:3' Wed Jan 1 01:02:03 CET 2020 $ TZ=Europe/Berlin date -d '2020-06-01 1:2:3' Mon Jun 1 01:02:03 CEST 2020 See? Could be CET or CEST. Japan doesn't seem to have daylight saving, so it'll be the same abbreviation throughout the year, but it can still change at some point. So always use a fixed date, like in your first test. 2) why is it CJT at all? $ TZ=Japan date -d '2009-10-04 22:23:00' +'%z %Z' +0900 JST it looks like it should be JST? ... 3) also, please test that CET/CEST example. Or, rather, MET/MEST/MST/MMT example - the data for these abbreviations are already present in mariadb_test_data_timezone.sql That is, add lines like SET @@time_zone='Europe/Moscow'; SELECT DATE_FORMAT('....', '%z %Z'); SELECT DATE_FORMAT('....', '%z %Z'); SELECT DATE_FORMAT('....', '%z %Z'); and with different dates it should return different abbreviations for the same time zone. > +# > +# output depends on system time information. Hence replacing > +# to avoid result diff test failures. > +# > +SET @@time_zone= default; > +SELECT DATE_FORMAT('2009-10+|-HH:MM ABC22:23:00', '%z %Z') AS current_timezone; 1) I don't understand what you're testing here :( 2) Anyway, note that you can put, say, --timezone=Europe/Berlin into the date_formats.opt file and the server will be restarted with this value in TZ. That is, you can force the system time zone to be a specific fixed value. > diff --git a/scripts/mariadb_test_data_timezone.sql b/scripts/mariadb_test_data_timezone.sql > index 8d07d413cef..83bb548478d 100644 > --- a/scripts/mariadb_test_data_timezone.sql > +++ b/scripts/mariadb_test_data_timezone.sql > @@ -14,8 +14,8 @@ > -- along with this program; if not, write to the Free Software > -- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA > > -INSERT INTO time_zone_name (Name, Time_Zone_id) ... > -INSERT INTO time_zone (Time_zone_id, Use_leap_seconds) ... > -INSERT INTO time_zone_transition (Time_zone_id, Transition_time, Transition_type_id) ... > -INSERT INTO time_zone_transition_type (Time_zone_id, Transition_type_id, `Offset`, Is_DST, Abbreviation) ... > -INSERT INTO time_zone_leap_second (Transition_time, Correction) ... > +INSERT INTO time_zone_name (Name, Time_Zone_id) VALUES ... ('India/Kolkata', 6); > +INSERT INTO time_zone (Time_zone_id, Use_leap_seconds) ... > +INSERT INTO time_zone_transition (Time_zone_id, Transition_time, Transition_type_id) ... > +INSERT INTO time_zone_transition_type (Time_zone_id, Transition_type_id, `Offset`, Is_DST, Abbreviation) ... > +INSERT INTO time_zone_leap_second (Transition_time, Correction) ... 1. Why? I didn't see where you use this new value in tests. 2. Is it really India/Kolkata? The official https://www.iana.org/time-zones seems to have Asia/Kolkata. > diff --git a/sql/item_timefunc.cc b/sql/item_timefunc.cc > index 26adc4eddaa..f8d3fe7dfe8 100644 > --- a/sql/item_timefunc.cc > +++ b/sql/item_timefunc.cc > @@ -468,20 +468,111 @@ static bool extract_date_time(THD *thd, DATE_TIME_FORMAT *format, > DBUG_RETURN(1); > } > > +/* > + > + Gets local time in MYSQL_TIME structure and tm struct. > + > +*/ > +void get_timezone_offset_information(THD *thd, Time_zone *tz, int &hr, > + int &min, char* abbrevation, > + bool &is_neg) > +{ > + if (!tz) > + return; > + > + if (tz->get_name()->length() == tz_SYSTEM_name.length() && > + !strncmp((const char*)(tz->get_name()->ptr()), > + (const char*)(tz_SYSTEM_name.ptr()), 6)) > + { > + struct tm tm_local_time; > + time_t time_sec= my_time(0); > + MYSQL_TIME local_TIME, gmt_TIME; > + ulonglong seconds_diff; > + ulong microsec_diff; > + > + /* Get local time in MYSQL_TIME struct. */ > + thd->variables.time_zone->gmt_sec_to_TIME(&local_TIME, > + thd->query_start()); > + /* Get GMT time in MYSQL_TIME struct. */ > + my_tz_UTC->gmt_sec_to_TIME(&gmt_TIME, thd->query_start()); > + > + /* We need above two MYSQL_TIME to get time difference. */ > + is_neg= calc_time_diff(&local_TIME, &gmt_TIME, 1, > + &seconds_diff, &microsec_diff); > + > + /* > + once we have time difference in seconds_diff, > + convert it into hr and min. > + */ > + hr= (int)(seconds_diff/3600L); > + int temp= (int)(seconds_diff%3600L); > + min= temp/60L; > + > + /* > + Get timezone abbrevation. It is stored in tm structure, > + so convert time to tm struct and copy it. > + */ 1) tm structure also has the offset, you don't need to calculate it manually. 2) you should use not my_time(0) but the value of the first DATE_FORMAT() argument > + localtime_r(&time_sec, &tm_local_time); > +# ifdef __USE_MISC > + int len= strlen((tm_local_time.tm_zone)); > + strncpy(abbrevation, (tm_local_time.tm_zone), len); > + abbrevation[len]= '\0'; > +# endif > + } > + else > + { > + /* > + Information for "named" timezone and offset can be obtained from their > + relevant class. So get the timezone information first. > + */ > + struct tz tmp; > + tz->get_timezone_information(&tmp); > + is_neg= tmp.is_behind; > + > + /* Get time in hour and min. */ > + time_t time_in_sec= abs(tmp.seconds_offset); > + hr= (int)(time_in_sec/3600L); > + int temp= (int)(time_in_sec%3600L); > + min= temp/60L; > + > + /* Copy the abbrevation. */ > + strncpy(abbrevation, (const char*)(tmp.abbrevation), > + strlen((const char*)(tmp.abbrevation))); > + abbrevation[strlen((const char*)(tmp.abbrevation))+1]='\0'; > + } > + > + return; > +} > + > > /** > Create a formatted date/time value in a string. > */ > > -static bool make_date_time(const String *format, const MYSQL_TIME *l_time, > - timestamp_type type, const MY_LOCALE *locale, > - String *str) > +static bool make_date_time(THD *thd, const String *format, > + const MYSQL_TIME *l_time, timestamp_type type, > + const MY_LOCALE *locale, String *str) > { > char intbuff[15]; > uint hours_i; > uint weekday; > ulong length; > const char *ptr, *end; > + int diff_hr=0, diff_min=0; > + bool curr_time_behind_utc= false; > + char abbrevation[8]; > + > + Time_zone* curr_timezone= my_tz_find(thd, > + thd->variables.time_zone->get_name()); > + > + /* > + precalculate local time and timezone information because > + %z or %Z maybe present in the format. > + */ > + memset(abbrevation, 0, sizeof(abbrevation)); > + get_timezone_offset_information(thd, curr_timezone, diff_hr, > + diff_min, abbrevation, > + curr_time_behind_utc); Just a thought. get_timezone_offset_information() doesn't seem to be very cheap. May be do it only when %z or %Z is requested? Like case 'z': if (!have_tz_info) { memset(abbrevation, 0, sizeof(abbrevation)); get_timezone_offset_information(thd, curr_timezone, diff_hr, diff_min, abbrevation, curr_time_behind_utc); have_tz_info= true; } Or better something with less code duplication :) > str->length(0); > > @@ -699,6 +790,31 @@ static bool make_date_time(const String *format, const MYSQL_TIME *l_time, > str->append_with_prefill(intbuff, length, 1, '0'); > break; > > + case 'z': > + { > + if (curr_time_behind_utc) > + str->append("-", 1, system_charset_info); > + else > + str->append("+", 1, system_charset_info); > + > + if (diff_hr/10 == 0) > + str->append("0", 1, system_charset_info); > + length= (uint) (int10_to_str(diff_hr, intbuff, 10) - intbuff); > + str->append_with_prefill(intbuff, length, 1, '0'); > + if (diff_min/10 == 0) > + str->append("0", 1, system_charset_info); > + length= (uint) (int10_to_str(diff_min, intbuff, 10) - intbuff); > + str->append_with_prefill(intbuff, length, 1, '0'); why do you append "0" manually, if you use append_with_prefill(), which, I suppose, does it automatically? or, if you add "0" manually, you should use simple append(). > + } > + break; > + > + case 'Z': > + { > + str->append(abbrevation, > + (uint) strlen(abbrevation), > + system_charset_info); > + } > + break; > default: > str->append(*ptr); > break; > @@ -1912,7 +2028,7 @@ String *Item_func_date_format::val_str(String *str) > > /* Create the result string */ > str->set_charset(collation.collation); > - if (!make_date_time(format, &l_time, > + if (!make_date_time(thd, format, &l_time, > is_time_format ? MYSQL_TIMESTAMP_TIME : > MYSQL_TIMESTAMP_DATE, > lc, str)) > diff --git a/sql/tztime.cc b/sql/tztime.cc > index e77a7529332..4662fa082e0 100644 > --- a/sql/tztime.cc > +++ b/sql/tztime.cc > @@ -55,15 +55,10 @@ > > /* > Now we don't use abbreviations in server but we will do this in future. > + Edit: Startd needing abbrevation in server as part of task MDEV-31684. typo > */ > -#if defined(TZINFO2SQL) || defined(TESTTIME) > -#define ABBR_ARE_USED > -#else > -#if !defined(DBUG_OFF) > -/* Let use abbreviations for debug purposes */ > -#undef ABBR_ARE_USED > -#define ABBR_ARE_USED > -#endif /* !defined(DBUG_OFF) */ > +#ifndef ABBR_ARE_USED > +#define ABBR_ARE_USED > #endif /* defined(TZINFO2SQL) || defined(TESTTIME) */ > > /* Structure describing local time type (e.g. Moscow summer time (MSD)) */ > @@ -1008,11 +1003,6 @@ TIME_to_gmt_sec(const MYSQL_TIME *t, const TIME_ZONE_INFO *sp, uint *error_code) > > #if !defined(TESTTIME) && !defined(TZINFO2SQL) > > -/* > - String with names of SYSTEM time zone. > -*/ > -static const String tz_SYSTEM_name("SYSTEM", 6, &my_charset_latin1); > - > > /* > Instance of this class represents local time zone used on this system > @@ -1031,6 +1021,7 @@ class Time_zone_system : public Time_zone > virtual my_time_t TIME_to_gmt_sec(const MYSQL_TIME *t, uint *error_code) const; > virtual void gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const; > virtual const String * get_name() const; > + virtual void get_timezone_information(struct tz* curr_tz) const; > }; > > > @@ -1097,6 +1088,12 @@ Time_zone_system::gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const > adjust_leap_second(tmp); > } > > +void > +Time_zone_system::get_timezone_information(struct tz* curr_tz) const > +{ this is weird. You have a special if (tz->get_name()->length() == tz_SYSTEM_name.length() && !strncmp((const char*)(tz->get_name()->ptr()), (const char*)(tz_SYSTEM_name.ptr()), 6)) to detect SYSTEM time zone by name in get_timezone_offset_information(). why not to put that logic here, in the dedicated method, like you did for other time zones? > + return; > +} > + > > /* > Get name of time zone > @@ -1128,6 +1125,7 @@ class Time_zone_utc : public Time_zone > uint *error_code) const; > virtual void gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const; > virtual const String * get_name() const; > + virtual void get_timezone_information(struct tz* curr_tz) const; > }; > > > @@ -1175,6 +1173,12 @@ Time_zone_utc::gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const > adjust_leap_second(tmp); > } > > +void > +Time_zone_utc::get_timezone_information(struct tz* curr_tz) const > +{ what is printed for UTC? Please, add a test for that. > + return; > +} > + > > /* > Get name of time zone > @@ -1210,6 +1214,7 @@ class Time_zone_db : public Time_zone > virtual my_time_t TIME_to_gmt_sec(const MYSQL_TIME *t, uint *error_code) const; > virtual void gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const; > virtual const String * get_name() const; > + virtual void get_timezone_information(struct tz* curr_tz) const; > private: > TIME_ZONE_INFO *tz_info; > const String *tz_name; > @@ -1280,6 +1285,16 @@ Time_zone_db::gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const > adjust_leap_second(tmp); > } > > +void > +Time_zone_db::get_timezone_information(struct tz* curr_tz) const > +{ > + curr_tz->seconds_offset= tz_info->revtis->rt_offset; > + curr_tz->is_behind= tz_info->revtis->rt_offset < 0 ? true : false; > + strncpy((char*)(curr_tz->abbrevation), (const char*)(tz_info->chars), > + strlen((const char*)(tz_info->chars))+1); > + curr_tz->abbrevation[strlen((const char*)(tz_info->chars))+1]='\0'; here and below (in Time_zone_offset) 1. why casts? 2. you use strncpy incorrectly, the third argument specifies the size of the buffer, like strncpy(curr_tz->abbrevation, tz_info->chars, sizeof(curr_tz->abbrevation)); but better use strmake, it'll do the zero-termination for you strmake(curr_tz->abbrevation, tz_info->chars, sizeof(curr_tz->abbrevation)-1); > +} > + > > /* > Get name of time zone > @@ -1309,6 +1324,7 @@ class Time_zone_offset : public Time_zone > uint *error_code) const; > virtual void gmt_sec_to_TIME(MYSQL_TIME *tmp, my_time_t t) const; > virtual const String * get_name() const; > + virtual void get_timezone_information(struct tz* curr_tz) const; > /* > This have to be public because we want to be able to access it from > my_offset_tzs_get_key() function > @@ -1434,6 +1450,16 @@ Time_zone_offset::get_name() const > return &name; > } > > +void > +Time_zone_offset::get_timezone_information(struct tz* curr_tz) const > +{ > + curr_tz->seconds_offset= offset; > + const char *name= get_name()->ptr(); > + curr_tz->is_behind= name[0] == '+' ? false : true; > + strncpy((char*)(curr_tz->abbrevation), name, this->get_name()->length()); > + curr_tz->abbrevation[this->get_name()->length()]= '\0'; > +} > + > > static Time_zone_utc tz_UTC; > static Time_zone_system tz_SYSTEM; > diff --git a/sql/tztime.h b/sql/tztime.h > index 6d8af62ecd4..57e4aa2021b 100644 > --- a/sql/tztime.h > +++ b/sql/tztime.h > @@ -26,6 +26,11 @@ > #include "sql_list.h" /* Sql_alloc */ > #include "sql_string.h" /* String */ > > +/* > + String with names of SYSTEM time zone. > +*/ > +static const String tz_SYSTEM_name("SYSTEM", 6, &my_charset_latin1); No, sorry. It's a .h file included in many .cc files. You'll have a separate `static const String` in every single source file that includes this header. > + > class THD; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 23b15d24eb3: cleanup: add Field::store_yesno
by Sergei Golubchik 31 Aug '23

31 Aug '23

Hi, Nikita, On Aug 31, Nikita Malyavin wrote: > revision-id: 23b15d24eb3 (mariadb-11.0.1-227-g23b15d24eb3) > parent(s): cb37c99dd87 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-08-30 13:18:52 +0400 > message: > > cleanup: add Field::store_yesno I agree, it's a good idea to have a helper for storing YES/NO in the field. But I think Field class itself is a wrong place for it. YES/NO is exclusively INFORMATION_SCHEMA thing and it feels kind of wrong to have it in the base class. I'd do a separate function, like store_yes_no(). Or something. Also, note that you missed few places where it's needed, like pos=(uchar*) ((field->flags & NOT_NULL_FLAG) ? "NO" : "YES"); table->field[6]->store((const char*) pos, strlen((const char*) pos), cs); or if (updatable_view) table->field[5]->store(STRING_WITH_LEN("YES"), cs); else table->field[5]->store(STRING_WITH_LEN("NO"), cs); both in sql_show.cc. There are more. Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: Client-server protocol improvement proposal: Connection redirection
by Otto Kekäläinen 28 Aug '23

28 Aug '23

Hi Pali! On Fri, 18 Aug 2023 at 11:47, Pali <pali(a)cpan.org> wrote: > CCing choroba (for Perl's DBD::MariaDB) and replying to developers@ ML. > > I have read proposal and the first thing which worries me is how the > proposal handles security aspect of the whole solution. > ... > If TLS is used then any protocol redirection information must happen > inside of the TLS tunnel. Thanks for the comments! Actually Daniel's submission has already this covered and the redirect works only if the connection is TLS protected: * https://github.com/MariaDB/server/pull/2681, server-side implementation * https://github.com/mariadb-corporation/mariadb-connector-c/pull/226, MariaDB Connector/C implementation

1 0

Re: [MariaDB commits] [PATCH 4/4] MDEV-31273: Precompute binlog checksums
by Marko Mäkelä 25 Aug '23

25 Aug '23

On Fri, Aug 25, 2023 at 10:16 AM Kristian Nielsen via commits <commits(a)lists.mariadb.org> wrote: > > Compute binlog checksums (when enabled) already when writing events > into the statement or transaction caches, where before it was done > when the caches are copied to the real binlog file. This moves the > checksum computation outside of holding LOCK_log, improving > scalabitily. > > At stmt/trx cache write time, the final end_log_pos values are not > known, so with this patch these will be set to 0. [...] > Checksums cannot be pre-computed when binlog encryption is enabled, as > encryption relies on correct end_log_pos to provide part of the > nonce/IV. Checksum pre-computation is also disabled for WSREP/Galera. Have you thought about a format change that would address these issues? In MDEV-14425 https://github.com/MariaDB/server/commit/685d958e38b825ad9829be311f26729ccc… I redesigned the InnoDB log file format in a way that allows checksums to be computed before the final position of the records (the log sequence number of LSN) is known. For encryption, an additional 64-bit nonce will be written; this currently is the LSN at the time the encryption was scheduled. Tablespace identifiers, page numbers and file names are not encrypted, because these were never fully encrypted in innodb_encrypt_tables=ON either. This allows the metadata to be used as initialization vectors, and also avoids the need for backup tools to know any encryption keys. Thanks to MDEV-27774, not only InnoDB computes checksums while not holding any mutex, but multiple threads can copy their mini-transactions to the shared log_sys.buf while holding a shared lock_sys.latch. An exclusive latch will only be needed during a log checkpoint when we need to guarantee that all changes up to a particular LSN have been written to the log buffer. This would not be possible without the format change that was implemented in MDEV-14425. > +/* Convenience macros since ulong is 32 or 64 bit depending on platform. */ > +#if SIZEOF_LONG == 4 > +#define my_atomic_storeul(P, D) my_atomic_store32((int32 volatile *)(P), (D)) In all currently supported MariaDB Server versions, C++11 is available. Is there a particular reason why you would not use std::atomic<size_t> here? Marko -- Marko Mäkelä, Lead Developer InnoDB MariaDB plc

2 3

Client-server protocol improvement proposal: Connection redirection
by Lenski, Daniel 18 Aug '23

18 Aug '23

Hi! I would like to propose a new feature in the MariaDB client-server protocol<https://mariadb.com/kb/en/clientserver-protocol>: application-layer redirection of client connections. We want the MariaDB server to be able to tell clients connecting to it, “Sorry, this server is unavailable. Connect to an alternate server instead.” This mechanism is inspired by HTTP 302 (“temporary redirect”) mechanism familiar to all developers of web applications, and is intended to have similar semantics and security properties, since these have now been widely deployed and tested for decades. I have submitted a minimal but viable implementation of this at: * https://github.com/MariaDB/server/pull/2681, server-side implementation * https://github.com/mariadb-corporation/mariadb-connector-c/pull/226, MariaDB Connector/C implementation I am seeking advice and opinions on these, and will proceed to finalize this in the following 1-2 months. In its current form, this implementation allows the MariaDB server administrator to set two variables: * SET GLOBAL SERVER_REDIRECT_MODE={ON,OFF} (the default is OFF) * SET GLOBAL SERVER_REDIRECT_TARGET='my-new-server.example.com' (or 192.168.0.123:3307, or new-server.com:3308, etc) When SERVER_REDIRECT_MODE is set to on, the server will stop accepting new TCP-based connections (unless they are to the extra port<https://mariadb.com/kb/en/thread-pool-in-mariadb/#configuring-the-extra-port>, which is intended for emergency administrative use), and it will instead respond to all clients with an error packet<https://mariadb.com/kb/en/err_packet/>: $ mariadb --host my-server.example.com ERROR 4189 (HY000): |Server is redirecting clients to 'my-new-server.example.com:3307'|my-new-server.example.com:3307 When an appropriately-updated client receives this error packet (error code 4196, message formatted as |Human readable message|<value of SERVER_REDIRECT_TARGET>), it will parse the redirection target out from the error message and attempt a new connection: $ updated-mariadb --host my-server.example.com -e 'select @@hostname' Got server redirect to 'my-new-server.example.com' (port 3307) +---------------------------+ | @@hostname | +---------------------------+ | my-new-server.example.com | +---------------------------+ The feature is gracefully backwards-compatible. Old clients that do not support redirection will fail and show the message in a form from which humans can deduce what happened. New clients that do support it will follow it to the new server. The feature is also designed to be extensible in a forwards-compatible way. The SERVER_REDIRECT_MODE system variable could be expanded to support additional redirection modes. For example, it could be augmented to redirect new connections only from specific IP subnets, or to direct the client to interpret the SERVER_REDIRECT_TARGET variable in a new form (such as a<https://dev.mysql.com/doc/connector-j/8.0/en/connector-j-reference-jdbc-url…> JDBC-compatible mysql:// URL<https://dev.mysql.com/doc/connector-j/8.0/en/connector-j-reference-jdbc-url…> as proposed in MDEV-15935<https://jira.mariadb.org/browse/MDEV-15935>). While researching and designing this connection redirection mechanism, I discovered or rediscovered a couple of major TLS-related security vulnerabilities in the MariaDB Connector/C client [1, 2]. These vulnerabilities affect all MariaDB users who care about the confidentiality and authenticity of their client/server connections, and they become even more severe if a connection redirection mechanism exists — because this entails the client trusting the server to send it instructions for creating a new network connection. Unless these are resolved, any connection redirection mechanism will be susceptible to hijacking and redirection to an attacker-controlled server even if the client thinks it is using TLS/SSL. What are the use cases for this connection redirection feature? Even in this simple, initial form, the feature is already useful for database administrators who want to gracefully shut down a MariaDB server (for example, to upgrade it), and to redirect new client connections to a replica or alternate server that can accept the traffic. Server-side connection redirection can be combined with existing MariaDB server features, such as SET GLOBAL READ_ONLY=ON and KILLing of idle or long-running connections, in order to build a full switchover solution for shunting traffic from one server to another as gracefully as possible. Currently, MariaDB database administrators tend to use proxies and wrappers (such as HikariCP) to redirect client connections from one MariaDB server to another. Supporting this redirection mechanism directly in the server, and in the client-server protocol, will allow many simple use cases of redirection without extra software, making MariaDB easier and simpler to manage for minimal downtime situations. Please post on this mailing list what you think about this suggestion. I am looking forward to polishing this feature and submitting it in a form that meets the quality requirements for new features from MariaDB core developers. Thank you! Daniel Lenski Amazon RDS MySQL/MariaDB engine team --- Notes on TLS vulnerabilities in MariaDB Connector/C: [1] CONC-648<https://jira.mariadb.org/browse/CONC-648>, reported by me on 5 June: when a Connector/C-based client is using connections secured by TLS/SSL, Connector/C will nevertheless receive plain-text, unencrypted error packets prior to the TLS handshake, and present them to the client in a way that’s indistinguishable from TLS-protected error packets sent by the server after the TLS handshake. This longstanding vulnerability is a consequence of the tangled and unsystematic way that the Connector/C codebase interleaves the MariaDB application-layer authentication protocol with the transport-layer security protocol. My standalone mariadb-connector-c PR#223<https://github.com/mariadb-corporation/mariadb-connector-c/pull/223> fixes the most straightforward manifestation of this vulnerability. Until this issue is resolved in Connector/C, it will be trivial for MITM attackers to inject plaintext connection redirection packets into MariaDB client/server connections, and thereby redirect clients to attacker-controlled servers despite those clients’ intention to use only TLS-secured connections. [2] MDEV-28634<https://jira.mariadb.org/browse/MDEV-28634>, reported almost 3 years ago by Geoff Montee: Connector/C clients which specify --ssl without --ssl-verify-server-cert will allow their connections to be silently downgraded to unencrypted, plaintext connections. This vulnerability can be used by on-path attackers to trivially downgrade most MariaDB client-server connections from TLS-encrypted to plaintext. My standalone mariadb-connector-c PR#224<https://github.com/mariadb-corporation/mariadb-connector-c/pull/224> fixes this. It is a backwards-incompatible change in that clients which specify --ssl (and thus clearly expect TLS-secured connections) will no longer succeed in connecting to servers which don’t support TLS. This should be considered a good thing: MariaDB Connector/C-based clients have been misleading their users about whether their connections are guaranteed to be encrypted, and will no longer do so. Essentially all software that purports to secure network connections with TLS have implemented similar changes, most of it about a decade ago when the pervasiveness of such attacks<https://en.wikipedia.org/wiki/Global_surveillance_disclosures_(2013%E2%80%9…> (by ISPs, censors, intelligenceagencies, etc) became very widely-known.

8 28

Re: [Commits] cb06612a9da: MDEV-31655: Parallel replication deadlock victim preference code errorneously removed
by Kristian Nielsen 15 Aug '23

15 Aug '23

Hi Marko, I'd like to ask you to review the below patch for MDEV-31655. https://jira.mariadb.org/browse/MDEV-31655 https://github.com/MariaDB/server/commit/cb06612a9da09a7981ada84768793f2ff3… This patch restores code in InnoDB that was originally part of the parallel replication implementation. It makes InnoDB choose the right victim when two transactions in parallel replication deadlock with each other. The code was erroneously removed as part of the VATS work in InnoDB. I saw that you later deprecated and eventually removed the VATS code. I saw that the InnoDB deadlock code is substantially changed in 10.6, so I will need to rework the patch in that version. But I wanted to ask your opinion about this 10.4 version first. In the patch, note in particular the modification to has_higher_priority(), which I think has the effect of choosing which of two waiting transactions to grant a lock to. I don't remember adding this myself, so it might have been added as part of VATS work. But it sounds appropriate; if T1 and T2 are both waiting for a lock, there is no point in granting it to T2 over T1, as then T2 will just be immediately killed and rolled back by replication. - Kristian. Kristian Nielsen via commits <commits(a)lists.mariadb.org> writes: > revision-id: cb06612a9da09a7981ada84768793f2ff3ef842c (mariadb-10.4.30-3-gcb06612a9da) > parent(s): dbe5c20b755b87f67d87990c3baabc866667e41b > author: Kristian Nielsen > committer: Kristian Nielsen > timestamp: 2023-07-11 00:31:29 +0200 > message: > > MDEV-31655: Parallel replication deadlock victim preference code errorneously removed > > Restore code to make InnoDB choose the second transaction as a deadlock > victim if two transactions deadlock that need to commit in-order for > parallel replication. This code was erroneously removed when VATS was > implemented in InnoDB. > > Also add a test case for InnoDB choosing the right deadlock victim. > > Signed-off-by: Kristian Nielsen <knielsen(a)knielsen-hq.org> > > --- > .../suite/binlog_encryption/rpl_parallel.result | 42 ++++++++++++- > mysql-test/suite/rpl/r/rpl_parallel.result | 42 ++++++++++++- > mysql-test/suite/rpl/t/rpl_parallel.test | 71 +++++++++++++++++++++- > sql/rpl_parallel.cc | 7 ++- > sql/sql_class.cc | 43 +++++++++++++ > storage/innobase/lock/lock0lock.cc | 12 ++++ > storage/innobase/trx/trx0trx.cc | 12 ++++ > 7 files changed, 225 insertions(+), 4 deletions(-) > > diff --git a/mysql-test/suite/binlog_encryption/rpl_parallel.result b/mysql-test/suite/binlog_encryption/rpl_parallel.result > index b75a66a634a..b24ff7ba53d 100644 > --- a/mysql-test/suite/binlog_encryption/rpl_parallel.result > +++ b/mysql-test/suite/binlog_encryption/rpl_parallel.result > @@ -2,6 +2,7 @@ include/master-slave.inc > [connection master] > connection server_2; > SET @old_parallel_threads=@@GLOBAL.slave_parallel_threads; > +SET @old_parallel_mode=@@GLOBAL.slave_parallel_mode; > SET GLOBAL slave_parallel_threads=10; > ERROR HY000: This operation cannot be performed as you have a running slave ''; run STOP SLAVE '' first > include/stop_slave.inc > @@ -1680,13 +1681,52 @@ a > 2000 > SELECT * FROM t2 WHERE a>=2000 ORDER BY a; > a > +MDEV-31655: Parallel replication deadlock victim preference code erroneously removed > +connection server_1; > +CREATE TABLE t7 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB; > +BEGIN; > +COMMIT; > +include/save_master_gtid.inc > +connection server_2; > +include/sync_with_master_gtid.inc > +include/stop_slave.inc > +set @@global.slave_parallel_threads= 5; > +set @@global.slave_parallel_mode= conservative; > +SET @old_dbug= @@GLOBAL.debug_dbug; > +SET GLOBAL debug_dbug= "+d,rpl_mdev31655_zero_retries"; > +connection master; > +SET @old_dbug= @@SESSION.debug_dbug; > +SET SESSION debug_dbug="+d,binlog_force_commit_id"; > +SET @commit_id= 1+1000; > +SET @commit_id= 2+1000; > +SET @commit_id= 3+1000; > +SET @commit_id= 4+1000; > +SET @commit_id= 5+1000; > +SET @commit_id= 6+1000; > +SET @commit_id= 7+1000; > +SET @commit_id= 8+1000; > +SET @commit_id= 9+1000; > +SET @commit_id= 10+1000; > +SET SESSION debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > +COUNT(*) SUM(a*100*b) > +10 225000 > +include/save_master_gtid.inc > +connection server_2; > +include/start_slave.inc > +include/sync_with_master_gtid.inc > +SET GLOBAL debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > +COUNT(*) SUM(a*100*b) > +10 225000 > connection server_2; > include/stop_slave.inc > SET GLOBAL slave_parallel_threads=@old_parallel_threads; > +SET GLOBAL slave_parallel_mode=@old_parallel_mode; > include/start_slave.inc > SET DEBUG_SYNC= 'RESET'; > connection server_1; > DROP function foo; > -DROP TABLE t1,t2,t3,t4,t5,t6; > +DROP TABLE t1,t2,t3,t4,t5,t6,t7; > SET DEBUG_SYNC= 'RESET'; > include/rpl_end.inc > diff --git a/mysql-test/suite/rpl/r/rpl_parallel.result b/mysql-test/suite/rpl/r/rpl_parallel.result > index 9b2e68d366e..ef89d954faa 100644 > --- a/mysql-test/suite/rpl/r/rpl_parallel.result > +++ b/mysql-test/suite/rpl/r/rpl_parallel.result > @@ -2,6 +2,7 @@ include/master-slave.inc > [connection master] > connection server_2; > SET @old_parallel_threads=@@GLOBAL.slave_parallel_threads; > +SET @old_parallel_mode=@@GLOBAL.slave_parallel_mode; > SET GLOBAL slave_parallel_threads=10; > ERROR HY000: This operation cannot be performed as you have a running slave ''; run STOP SLAVE '' first > include/stop_slave.inc > @@ -1679,13 +1680,52 @@ a > 2000 > SELECT * FROM t2 WHERE a>=2000 ORDER BY a; > a > +MDEV-31655: Parallel replication deadlock victim preference code erroneously removed > +connection server_1; > +CREATE TABLE t7 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB; > +BEGIN; > +COMMIT; > +include/save_master_gtid.inc > +connection server_2; > +include/sync_with_master_gtid.inc > +include/stop_slave.inc > +set @@global.slave_parallel_threads= 5; > +set @@global.slave_parallel_mode= conservative; > +SET @old_dbug= @@GLOBAL.debug_dbug; > +SET GLOBAL debug_dbug= "+d,rpl_mdev31655_zero_retries"; > +connection master; > +SET @old_dbug= @@SESSION.debug_dbug; > +SET SESSION debug_dbug="+d,binlog_force_commit_id"; > +SET @commit_id= 1+1000; > +SET @commit_id= 2+1000; > +SET @commit_id= 3+1000; > +SET @commit_id= 4+1000; > +SET @commit_id= 5+1000; > +SET @commit_id= 6+1000; > +SET @commit_id= 7+1000; > +SET @commit_id= 8+1000; > +SET @commit_id= 9+1000; > +SET @commit_id= 10+1000; > +SET SESSION debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > +COUNT(*) SUM(a*100*b) > +10 225000 > +include/save_master_gtid.inc > +connection server_2; > +include/start_slave.inc > +include/sync_with_master_gtid.inc > +SET GLOBAL debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > +COUNT(*) SUM(a*100*b) > +10 225000 > connection server_2; > include/stop_slave.inc > SET GLOBAL slave_parallel_threads=@old_parallel_threads; > +SET GLOBAL slave_parallel_mode=@old_parallel_mode; > include/start_slave.inc > SET DEBUG_SYNC= 'RESET'; > connection server_1; > DROP function foo; > -DROP TABLE t1,t2,t3,t4,t5,t6; > +DROP TABLE t1,t2,t3,t4,t5,t6,t7; > SET DEBUG_SYNC= 'RESET'; > include/rpl_end.inc > diff --git a/mysql-test/suite/rpl/t/rpl_parallel.test b/mysql-test/suite/rpl/t/rpl_parallel.test > index 9ba7a30f2eb..d43cec4df34 100644 > --- a/mysql-test/suite/rpl/t/rpl_parallel.test > +++ b/mysql-test/suite/rpl/t/rpl_parallel.test > @@ -13,6 +13,7 @@ > > --connection server_2 > SET @old_parallel_threads=@@GLOBAL.slave_parallel_threads; > +SET @old_parallel_mode=@@GLOBAL.slave_parallel_mode; > --error ER_SLAVE_MUST_STOP > SET GLOBAL slave_parallel_threads=10; > --source include/stop_slave.inc > @@ -2203,16 +2204,84 @@ SELECT * FROM t1 WHERE a>=2000 ORDER BY a; > SELECT * FROM t2 WHERE a>=2000 ORDER BY a; > > > +--echo MDEV-31655: Parallel replication deadlock victim preference code erroneously removed > +# The problem was that InnoDB would choose the wrong deadlock victim. > +# Create a lot of transactions that can cause deadlocks, and use error > +# injection to check that the first transactions in each group is never > +# selected as deadlock victim. > +--let $rows= 10 > +--let $transactions= 5 > +--let $gcos= 10 > + > +--connection server_1 > +CREATE TABLE t7 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB; > +BEGIN; > +--disable_query_log > +--let $i= 0 > +while ($i < 10) { > + eval INSERT INTO t7 VALUES ($i, 0); > + inc $i; > +} > +--enable_query_log > +COMMIT; > +--source include/save_master_gtid.inc > + > +--connection server_2 > +--source include/sync_with_master_gtid.inc > +--source include/stop_slave.inc > +eval set @@global.slave_parallel_threads= $transactions; > +set @@global.slave_parallel_mode= conservative; > +SET @old_dbug= @@GLOBAL.debug_dbug; > +# This error injection will allow no retries for GTIDs divisible by 1000. > +SET GLOBAL debug_dbug= "+d,rpl_mdev31655_zero_retries"; > + > +--connection master > +SET @old_dbug= @@SESSION.debug_dbug; > +SET SESSION debug_dbug="+d,binlog_force_commit_id"; > + > +--let $j= 1 > +while ($j <= $gcos) { > + eval SET @commit_id= $j+1000; > + --let $i= 0 > + while ($i < $transactions) { > + --disable_query_log > + eval SET SESSION gtid_seq_no= 1000 + 1000*$j + $i; > + BEGIN; > + --let $k= 0 > + while ($k < $rows) { > + eval UPDATE t7 SET b=b+1 WHERE a=(($i+$k) MOD $rows); > + inc $k; > + } > + COMMIT; > + --enable_query_log > + inc $i; > + } > + inc $j; > +} > + > +SET SESSION debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > + > +--source include/save_master_gtid.inc > + > +--connection server_2 > +--source include/start_slave.inc > +--source include/sync_with_master_gtid.inc > +SET GLOBAL debug_dbug= @old_dbug; > +SELECT COUNT(*), SUM(a*100*b) FROM t7; > + > + > # Clean up. > --connection server_2 > --source include/stop_slave.inc > SET GLOBAL slave_parallel_threads=@old_parallel_threads; > +SET GLOBAL slave_parallel_mode=@old_parallel_mode; > --source include/start_slave.inc > SET DEBUG_SYNC= 'RESET'; > > --connection server_1 > DROP function foo; > -DROP TABLE t1,t2,t3,t4,t5,t6; > +DROP TABLE t1,t2,t3,t4,t5,t6,t7; > SET DEBUG_SYNC= 'RESET'; > > --source include/rpl_end.inc > diff --git a/sql/rpl_parallel.cc b/sql/rpl_parallel.cc > index b550315d69f..1aeb1257c4a 100644 > --- a/sql/rpl_parallel.cc > +++ b/sql/rpl_parallel.cc > @@ -1385,8 +1385,13 @@ handle_rpl_parallel_thread(void *arg) > err= dbug_simulate_tmp_error(rgi, thd);); > if (unlikely(err)) > { > + ulong max_retries= slave_trans_retries; > convert_kill_to_deadlock_error(rgi); > - if (has_temporary_error(thd) && slave_trans_retries > 0) > + DBUG_EXECUTE_IF("rpl_mdev31655_zero_retries", > + if ((rgi->current_gtid.seq_no % 1000) == 0) > + max_retries= 0; > + ); > + if (has_temporary_error(thd) && max_retries > 0) > err= retry_event_group(rgi, rpt, qev); > } > } > diff --git a/sql/sql_class.cc b/sql/sql_class.cc > index 8ed3f8a9c5e..e6ed7ca1cc4 100644 > --- a/sql/sql_class.cc > +++ b/sql/sql_class.cc > @@ -5247,6 +5247,49 @@ thd_need_ordering_with(const MYSQL_THD thd, const MYSQL_THD other_thd) > return 0; > } > > + > +/* > + If the storage engine detects a deadlock, and needs to choose a victim > + transaction to roll back, it can call this function to ask the upper > + server layer for which of two possible transactions is prefered to be > + aborted and rolled back. > + > + In parallel replication, if two transactions are running in parallel and > + one is fixed to commit before the other, then the one that commits later > + will be prefered as the victim - chosing the early transaction as a victim > + will not resolve the deadlock anyway, as the later transaction still needs > + to wait for the earlier to commit. > + > + The return value is -1 if the first transaction is prefered as a deadlock > + victim, 1 if the second transaction is prefered, or 0 for no preference (in > + which case the storage engine can make the choice as it prefers). > +*/ > +extern "C" int > +thd_deadlock_victim_preference(const MYSQL_THD thd1, const MYSQL_THD thd2) > +{ > + rpl_group_info *rgi1, *rgi2; > + > + if (!thd1 || !thd2) > + return 0; > + > + /* > + If the transactions are participating in the same replication domain in > + parallel replication, then request to select the one that will commit > + later (in the fixed commit order from the master) as the deadlock victim. > + */ > + rgi1= thd1->rgi_slave; > + rgi2= thd2->rgi_slave; > + if (rgi1 && rgi2 && > + rgi1->is_parallel_exec && > + rgi1->rli == rgi2->rli && > + rgi1->current_gtid.domain_id == rgi2->current_gtid.domain_id) > + return rgi1->gtid_sub_id < rgi2->gtid_sub_id ? 1 : -1; > + > + /* No preferences, let the storage engine decide. */ > + return 0; > +} > + > + > extern "C" int thd_non_transactional_update(const MYSQL_THD thd) > { > return(thd->transaction.all.modified_non_trans_table); > diff --git a/storage/innobase/lock/lock0lock.cc b/storage/innobase/lock/lock0lock.cc > index b1cf2152cd6..469d03eaa06 100644 > --- a/storage/innobase/lock/lock0lock.cc > +++ b/storage/innobase/lock/lock0lock.cc > @@ -49,6 +49,8 @@ Created 5/7/1996 Heikki Tuuri > #include <mysql/service_wsrep.h> > #endif /* WITH_WSREP */ > > +extern "C" int thd_deadlock_victim_preference(const MYSQL_THD thd1, const MYSQL_THD thd2); > + > /** Lock scheduling algorithm */ > ulong innodb_lock_schedule_algorithm; > > @@ -1538,6 +1540,16 @@ static bool has_higher_priority(lock_t *lock1, lock_t *lock2) > } else if (!lock_get_wait(lock2)) { > return false; > } > + // Ask the upper server layer if any of the two trx should be prefered. > + int preference = thd_deadlock_victim_preference(lock1->trx->mysql_thd, > + lock2->trx->mysql_thd); > + if (preference == -1) { > + // lock1 is preferred as a victim, so lock2 has higher priority > + return false; > + } else if (preference == 1) { > + // lock2 is preferred as a victim, so lock1 has higher priority > + return true; > + } > return lock1->trx->start_time_micro <= lock2->trx->start_time_micro; > } > > diff --git a/storage/innobase/trx/trx0trx.cc b/storage/innobase/trx/trx0trx.cc > index 7cd95878b0c..0771d764fb6 100644 > --- a/storage/innobase/trx/trx0trx.cc > +++ b/storage/innobase/trx/trx0trx.cc > @@ -52,6 +52,9 @@ Created 3/26/1996 Heikki Tuuri > #include <set> > #include <new> > > +extern "C" > +int thd_deadlock_victim_preference(const MYSQL_THD thd1, const MYSQL_THD thd2); > + > /** The bit pattern corresponding to TRX_ID_MAX */ > const byte trx_id_max_bytes[8] = { > 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff > @@ -1906,6 +1909,15 @@ trx_weight_ge( > { > ibool a_notrans_edit; > ibool b_notrans_edit; > + int pref; > + > + /* First ask the upper server layer if it has any preference for which > + to prefer as a deadlock victim. */ > + pref= thd_deadlock_victim_preference(a->mysql_thd, b->mysql_thd); > + if (pref < 0) > + return FALSE; > + else if (pref > 0) > + return TRUE; > > /* If mysql_thd is NULL for a transaction we assume that it has > not edited non-transactional tables. */ > _______________________________________________ > commits mailing list -- commits(a)lists.mariadb.org > To unsubscribe send an email to commits-leave(a)lists.mariadb.org

2 7

Re: 9c52eede2c8: MDEV-31646 Online alter applies binlog cache limit to cache writes
by Sergei Golubchik 15 Aug '23

15 Aug '23

Hi, Nikita, On Aug 14, Nikita Malyavin wrote: > revision-id: 9c52eede2c8 (mariadb-11.0.1-175-g9c52eede2c8) > parent(s): 209c894b431 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-08-11 16:29:37 +0400 > message: > > MDEV-31646 Online alter applies binlog cache limit to cache writes > > 1. Make online disk writes unlimited, same as filesort does. > 2. Make proper error handling -- in 32-bit build IO_CACHE capacity limit is > 4GB, so it is quite possible to overfill there. > 3. Event_log::write_cache complicated with event reparsing, and as it was > proven by QA, contains some mistakes. Rewrite introbuce a simpler and much > faster version, not featuring reparsing and therefore copying a whole > buffer at once. This also disables checksums and crypto. > > As a result, online alter is untied of several binlog variables, which was > a second aim of this patch. > > diff --git a/mysql-test/main/alter_table_online_debug.result b/mysql-test/main/alter_table_online_debug.result > index 021267a15f5..3a615372ed5 100644 > --- a/mysql-test/main/alter_table_online_debug.result > +++ b/mysql-test/main/alter_table_online_debug.result > @@ -409,23 +409,23 @@ set debug_sync= 'now SIGNAL start_replication'; > set debug_sync= 'now WAIT_FOR applied'; > select stage, progress from INFORMATION_SCHEMA.PROCESSLIST where id = @con; > stage progress > -3 51.220 > +3 53.390 why is that? > set debug_sync= 'now SIGNAL proceed WAIT_FOR applied'; > select stage, progress from INFORMATION_SCHEMA.PROCESSLIST where id = @con; > stage progress > diff --git a/sql/log.cc b/sql/log.cc > index d9e81700463..8c57487d57e 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -2286,13 +2286,21 @@ int binlog_log_row_online_alter(TABLE* table, const uchar *before_record, > MY_BITMAP *old_rpl_write_set= table->rpl_write_set; > table->rpl_write_set= &table->s->all_set; > > + table->online_alter_cache->store_prev_position(); > int error= (*log_func)(thd, table, table->s->online_alter_binlog, > - table->online_alter_cache, true, > + table->online_alter_cache, > + table->file->has_transactions_and_rollback(), > before_record, after_record); > > table->rpl_write_set= old_rpl_write_set; > > - return unlikely(error) ? HA_ERR_RBR_LOGGING_FAILED : 0; > + if (unlikely(error)) > + { > + table->online_alter_cache->restore_prev_position(); what is this for? > + return HA_ERR_RBR_LOGGING_FAILED; > + } > + > + return 0; > } > > static void > @@ -7748,7 +7763,7 @@ static int binlog_online_alter_end_trans(THD *thd, bool all, bool commit) > { > DBUG_ASSERT(cache.cache_log.type != READ_CACHE); > mysql_mutex_lock(binlog->get_log_lock()); > - error= binlog->write_cache(thd, &cache.cache_log); > + error= binlog->write_cache_raw(thd, &cache.cache_log); > mysql_mutex_unlock(binlog->get_log_lock()); > } > } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: ca64ddcc709: MDEV-31646 Online alter applies binlog cache limit to cache writes
by Sergei Golubchik 10 Aug '23

10 Aug '23

Hi, Nikita, This is a review of all the three commits On Jul 20, Nikita Malyavin wrote: > revision-id: ca64ddcc709 (mariadb-11.0.1-144-gca64ddcc709) > parent(s): f34c7419cb8 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-07-19 02:31:32 +0400 > message: > > MDEV-31646 Online alter applies binlog cache limit to cache writes > diff --git a/sql/log.cc b/sql/log.cc > index afd0643fe82..7466d63d325 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -2291,8 +2294,14 @@ int binlog_log_row_online_alter(TABLE* table, const uchar *before_record, > before_record, after_record); > > table->rpl_write_set= old_rpl_write_set; > + thd->pop_internal_handler(); > + > + if (unlikely(error)) > + push_warning_printf(thd, Sql_condition::WARN_LEVEL_NOTE, ER_DISK_FULL, > + "Broken online alter log. " > + "ALTER TABLE will finish with error."); I don't think so. Why would a DML statement get a warning about the concurrently running ALTER TABLE? It didn't do anything wrong, I don't think there should be a warning here Like, someone is running a series of DML statements, or the same statement many times. At some point another connection runs ALTER TABLE and suddenly those perfectly good DML statements start spewing out warnings. ALTER ends and all is clear again - looks wrong to me > > - return unlikely(error) ? HA_ERR_RBR_LOGGING_FAILED : 0; > + return 0; > } > > static void > @@ -2592,7 +2601,7 @@ void Event_log::set_write_error(THD *thd, bool is_transactional) > DBUG_VOID_RETURN; > } > > -bool Event_log::check_write_error(THD *thd) > +bool MYSQL_BIN_LOG::check_write_error(THD *thd) What's the difference? it's a static method anyway, FWIW it could be outside of any class. What's the point of moving it from one class to another? > { > DBUG_ENTER("MYSQL_BIN_LOG::check_write_error"); > > @@ -3806,9 +3815,9 @@ bool MYSQL_BIN_LOG::open_index_file(const char *index_file_name_arg, > } > > > -bool Event_log::open(enum cache_type io_cache_type_arg) > +bool Event_log::open(enum cache_type io_cache_type_arg, size_t buffer_size) > { > - bool error= init_io_cache(&log_file, -1, LOG_BIN_IO_SIZE, io_cache_type_arg, > + bool error= init_io_cache(&log_file, -1, buffer_size, io_cache_type_arg, why? it's only used by online alter, where would you need a different buffer size? only in your DBUG_EXECUTE_IF? > 0, 0, MYF(MY_WME | MY_NABP | MY_WAIT_IF_FULL)); > > log_state= LOG_OPENED; > @@ -6544,7 +6562,8 @@ Event_log::flush_and_set_pending_rows_event(THD *thd, Rows_log_event* event, > if (writer.write(pending)) > { > set_write_error(thd, is_transactional); > - if (check_write_error(thd) && cache_data && > + if (dynamic_cast<MYSQL_BIN_LOG*>(this) && why? > + MYSQL_BIN_LOG::check_write_error(thd) && cache_data && > stmt_has_updated_non_trans_table(thd)) > cache_data->set_incident(); > delete pending; > @@ -7721,8 +7749,11 @@ static int binlog_online_alter_end_trans(THD *thd, bool all, bool commit) > { > DBUG_ASSERT(cache.cache_log.type != READ_CACHE); > mysql_mutex_lock(binlog->get_log_lock()); > - error= binlog->write_cache(thd, &cache.cache_log); > + error= binlog->write_cache_raw(thd, &cache.cache_log); > mysql_mutex_unlock(binlog->get_log_lock()); > + > + if (unlikely(error)) > + binlog->set_write_error(my_errno); shouldn't binlog->write_cache() call set_write_error() internally? why would the caller have to do it? > } > } > else if (!commit) // rollback > @@ -7736,14 +7767,14 @@ static int binlog_online_alter_end_trans(THD *thd, bool all, bool commit) > cache.store_prev_position(); > } > > + thd->pop_internal_handler(); > > if (error) > { > - my_error(ER_ERROR_ON_WRITE, MYF(ME_ERROR_LOG), > - binlog->get_name(), errno); > - binlog_online_alter_cleanup(thd->online_alter_cache_list, > - is_ending_transaction); > - DBUG_RETURN(error); > + push_warning_printf(thd, Sql_state_errno_level::WARN_LEVEL_WARN, > + ER_ERROR_ON_WRITE, ER_THD(thd, ER_ERROR_ON_WRITE), > + binlog->get_name(), errno); > + DBUG_ASSERT(binlog->get_write_error()); same comment as above about not showing ALTER problems in concurrent threads > } > } > > diff --git a/sql/log.h b/sql/log.h > index 83e323cb6fb..39512b01991 100644 > --- a/sql/log.h > +++ b/sql/log.h > @@ -444,17 +445,20 @@ class Event_log: public MYSQL_LOG > */ > class Cache_flip_event_log: public Event_log { > IO_CACHE alt_buf; > +IF_DBUG(public:,) yuck I don't see you using online_write_error or ref_count in DBUG macros > IO_CACHE *current, *alt; > std::atomic<uint> ref_count; > + std::atomic<int> online_write_error; > public: > > Cache_flip_event_log() : Event_log(), alt_buf{}, > - current(&log_file), alt(&alt_buf), ref_count(1) {} > - bool open(enum cache_type io_cache_type_arg) > + current(&log_file), alt(&alt_buf), ref_count(1), > + online_write_error(0) {} > + bool open(size_t buffer_size) looks like Cache_flip_event_log::open() doesn't need any arguments at all > { > log_file.dir= mysql_tmpdir; > alt_buf.dir= log_file.dir; > - bool res= Event_log::open(io_cache_type_arg); > + bool res= Event_log::open(WRITE_CACHE); > if (res) > return res; > > diff --git a/sql/log_event.cc b/sql/log_event.cc > index 10180a95d99..0c4c284a40f 100644 > --- a/sql/log_event.cc > +++ b/sql/log_event.cc > @@ -761,16 +761,21 @@ Log_event::Log_event(const uchar *buf, > > int Log_event::read_log_event(IO_CACHE* file, String* packet, > const Format_description_log_event *fdle, > - enum enum_binlog_checksum_alg checksum_alg_arg) > + enum enum_binlog_checksum_alg checksum_alg_arg, > + size_t max_allowed_packet) > { > ulong data_len; > char buf[LOG_EVENT_MINIMAL_HEADER_LEN]; > uchar ev_offset= packet->length(); > #if !defined(MYSQL_CLIENT) > - THD *thd=current_thd; > - ulong max_allowed_packet= thd ? thd->slave_thread ? slave_max_allowed_packet > - : thd->variables.max_allowed_packet > - : ~(uint)0; > + if (max_allowed_packet == 0) may be better pass THD as an argument? or even correct value of max_allowed_packet. As ulong max_allowed_packet(THD *thd) { return thd ? ... : ~(uint)0; } on the second thought, do you need to ignore max_allowed_packet here for online ALTER? Is there no event size limit when the event is written into binlog? > + { > + THD *thd=current_thd; > + max_allowed_packet= thd ? thd->slave_thread > + ? slave_max_allowed_packet > + : thd->variables.max_allowed_packet > + : ~(uint)0; > + } > #endif > DBUG_ENTER("Log_event::read_log_event(IO_CACHE*,String*...)"); > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 7

Re: 88a77860a8b: MDEV-31804 Assertion `thd->m_transaction_psi == __null' fails
by Sergei Golubchik 09 Aug '23

09 Aug '23

Hi, Nikita, On Aug 08, Nikita Malyavin wrote: > revision-id: 88a77860a8b (mariadb-11.0.1-194-g88a77860a8b) > parent(s): 73a677f0306 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-08-07 22:45:11 +0400 > message: > > MDEV-31804 Assertion `thd->m_transaction_psi == __null' fails > > ... upon replicating online ALTER > > When an online event is applied and slave_exec_mode is idempotent, > Write_rows_log_event::do_before_row_operations had reset > thd->lex->sql_command to SQLCOM_REPLACE. > > This led to that a statement was detected as a row-type during > binlogging, and was logged as not standalone. > > So the corresponding Gtid_log_event, when applied on replica, did not > exit early and created a new PSI transaction. Hence the difference > with non-online ALTER. Nice find. I vaguely believe I've seen in one of the commits I've reviewed a change that makes every row event set thd->lex->sql_command to the corresponding SQLCOM_INSERT/UPDATE/DELETE. But be it was in an unpushed commit. Anyway, I think it's a bit risky to assume that row events never change thd->lex->sql_command (and never will). Perhaps you should simply restore it after the applying part? Like --- a/sql/sql_table.cc +++ b/sql/sql_table.cc @@ -12181,6 +12181,7 @@ copy_data_between_tables(THD *thd, TABLE *from, TABLE *> if (error) from->s->tdc->flush_unused(1); // to free the binlog to->pos_in_table_list= NULL; // Safety + thd->lex->sql_command= SQLCOM_ALTER_TABLE; } else if (online) // error was on copy stage { Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 5

Re: 63c93f89af9: MDEV-31804 Assertion `thd->m_transaction_psi == __null' fails
by Sergei Golubchik 07 Aug '23

07 Aug '23

Hi, Nikita, On Aug 07, Nikita Malyavin wrote: > revision-id: 63c93f89af9 (mariadb-11.0.1-192-g63c93f89af9) > parent(s): a75ba4caa51 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-08-05 20:17:41 +0400 > message: > > MDEV-31804 Assertion `thd->m_transaction_psi == __null' fails ... > index 1c8d56cafd1..66b2b1ee4cc 100644 > --- a/sql/handler.cc > +++ b/sql/handler.cc > @@ -1459,7 +1459,7 @@ void trans_register_ha(THD *thd, bool all, handlerton *ht_arg, ulonglong trxid) > Do not register transactions in which binary log is the only participating > transactional storage engine. > */ > - if (thd->m_transaction_psi == NULL && ht_arg->db_type != DB_TYPE_BINLOG) > + if (thd->m_transaction_psi == NULL && ht_arg != binlog_hton) it would've been better to fix it in 10.4, but if it doesn't show before online alter then ok > { > thd->m_transaction_psi= MYSQL_START_TRANSACTION(&thd->m_transaction_state, > thd->get_xid(), trxid, thd->tx_isolation, thd->tx_read_only, > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > index 88302096272..944807ea422 100644 > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -65,6 +65,7 @@ > #include "rpl_rli.h" > #include "log.h" > #include "sql_debug.h" > +#include "mysql/psi/mysql_transaction.h" > > #ifdef _WIN32 > #include <io.h> > @@ -11653,6 +11654,9 @@ bool mysql_trans_commit_alter_copy_data(THD *thd) > > DEBUG_SYNC(thd, "alter_table_copy_trans_commit"); > > + MYSQL_COMMIT_TRANSACTION(thd->m_transaction_psi); > + thd->m_transaction_psi= NULL; I don't understand it. ha_enable_transaction() commits the transaction it calls ha_commit_trans(). And ha_commit_trans() does the above psi manipulations. Why do you need to do that explicitly? Also, as a cleanup, I suspect, you can remove explicit commit that follows ha_enable_transaction() in mysql_trans_commit_alter_copy_data(). Seems to be redundant. > if (ha_enable_transaction(thd, TRUE)) > DBUG_RETURN(TRUE); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 2

Re: b508107c7a3: MDEV-31631 Adding auto-increment to table with history online misbehaves
by Sergei Golubchik 04 Aug '23

04 Aug '23

Hi, Nikita, Looks good. One comment below: On Aug 04, Nikita Malyavin wrote: > revision-id: b508107c7a3 (mariadb-11.0.1-189-gb508107c7a3) > parent(s): e9a75c98781 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-07-29 00:13:48 +0400 > message: > > MDEV-31631 Adding auto-increment to table with history online misbehaves > > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > index eabee13e8d3..3a70ed672a6 100644 > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -9953,6 +9953,37 @@ const char *online_alter_check_supported(const THD *thd, > } > } > > + for (auto &c: alter_info->create_list) > + { > + *online= c.field || !(c.flags & AUTO_INCREMENT_FLAG); > + if (!*online) > + return "ADD COLUMN ... AUTO_INCREMENT"; > + > + auto *def= c.default_value; > + *online= !(def && def->flags & VCOL_NEXTVAL > + // either it's a new field, or a NULL -> NOT NULL change > + && (!c.field || (!(c.field->flags & NOT_NULL_FLAG) > + && (c.flags & NOT_NULL_FLAG)))); > + if (!*online) > + { > + const char *fmt= ER_THD(thd, ER_GENERATED_COLUMN_FUNCTION_IS_NOT_ALLOWED); > + > + Item::vcol_func_processor_result res; > + IF_DBUG(bool ret=,) > + def->expr->walk(&Item::check_vcol_func_processor, 0, &res); I don't think you need to walk here, res.name is always "nextval()" isn't it? > + DBUG_ASSERT(!ret); > + DBUG_ASSERT(res.errors & VCOL_NEXTVAL); > + > + LEX_CSTRING dflt{STRING_WITH_LEN("DEFAULT")}; > + size_t len= strlen(fmt) + strlen(res.name) + c.field_name.length > + + dflt.length; > + char *resp= (char*)thd->alloc(len); > + // expression %s cannot be used in the %s clause of %`s > + my_snprintf(resp, len, fmt, res.name, dflt.str, c.field_name.str); > + return resp; > + } > + } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Running msqld with thread sanitizer
by Thejaka Amila J Kanewala 04 Aug '23

04 Aug '23

Hello, I am trying to build and run a thread sanitized mysqld binary. In the CMake configuration, I see the WITH_ASAN option, however, I do not see an option to build mysqld with the thread sanitizer. After making some custom configurations, I was able to build a mysqld binary with the thread sanitizer (by linking static thread sanitizer), however, when trying to run it, I got the following error: "ThreadSanitizer: sync allocator overflow (65536*1024). Dying." This error seems to be coming from the tsan library and seems this error is caused by too many synchronizations [1]. Reducing the number of threads also didn't help me with this error. So my questions to you experts: 1. Do you run mysqld with the thread sanitizer ? 2. If yes, could you please share the details. (e.g., CMake configurations, compiler/linker flags etc.) 3. If not, any suggestions to overcome the above error ? [1] https://groups.google.com/g/thread-sanitizer/c/AX5twD_IPP4 Thank you. Best Regards, Thejaka Amila Kanewala, PhD https://github.com/thejkane/agm

2 3

Re: 44b4a3e64fa: MDEV-31777 ER_GET_ERRNO upon online alter on CONNECT table
by Sergei Golubchik 03 Aug '23

03 Aug '23

Hi, Nikita, Generally, please, prefer table flags over new virtual function. E.g. *online= new_table->file->ha_table_flags() & HA_CAN_ONLINE_ALTER; But in this particular case there's no reason for a engine to not support online alter, it's just CONNECT being very weird. Like you wrote, it doesn't support REPLACE either and we didn't introduce a flag for that. This should be solved inside CONNECT. E.g it can throw an error just like it does for REPLACE. It can upgrade the lock: --- a/storage/connect/ha_connect.cc +++ b/storage/connect/ha_connect.cc @@ -5125,10 +5125,12 @@ @see get_lock_data() in lock.cc */ -THR_LOCK_DATA **ha_connect::store_lock(THD *, +THR_LOCK_DATA **ha_connect::store_lock(THD *thd, THR_LOCK_DATA **to, enum thr_lock_type lock_type) { + if (thd_sql_command(thd) == SQLCOM_ALTER_TABLE && lock_type == TL_READ) + lock_type= TL_READ_NO_INSERT; if (lock_type != TL_IGNORE && lock.type == TL_UNLOCK) lock.type=lock_type; *to++= &lock; or it can implement reading and inserting to the same table :) Either way, I don't think we need to extend the storage engine API for this specific case. Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org On Aug 03, Nikita Malyavin wrote: > revision-id: 44b4a3e64fa (mariadb-11.0.1-190-g44b4a3e64fa) > parent(s): b508107c7a3 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-07-29 00:13:48 +0400 > message: > > MDEV-31777 ER_GET_ERRNO upon online alter on CONNECT table > > Forbid Online for CONNECT. > > diff --git a/sql/handler.h b/sql/handler.h > index 480c79017ab..4a1363d2e93 100644 > --- a/sql/handler.h > +++ b/sql/handler.h > @@ -4911,6 +4911,8 @@ class handler :public Sql_alloc > bool commit); > > > + virtual bool is_online_alter_supported() const { return true; } > + > protected:

1 0

Re: dbe5c20b755: MDEV-31482: Lock wait timeout with INSERT-SELECT, autoinc, and statement-based replication
by Kristian Nielsen 03 Aug '23

03 Aug '23

Hi Marko, I'd like to ask your opinion/review of the below patch for MDEV-31482: https://github.com/MariaDB/server/commit/dbe5c20b755b87f67d87990c3baabc8666… https://jira.mariadb.org/browse/MDEV-31482 As you know, in-order parallel replication has InnoDB report lock waits. If T1 must replicate before T2, and InnoDB reports that T1 is waiting for a lock held by T2, then T2 will be killed (and retried later) so that T1 can replicate. As an optimisation, auto-increment lock waits where _not_ reported, but this turns out to be a bug. This patch just removes the exception so that such waits are reported like other waits. My only concern is the possible performance impact of this. I'm hoping it will not be very big, because in many cases the auto-increment lock is not taken for replicated transactions? My understanding is: - It is not taken for INSERT INTO t VALUES (...) (and maybe not for multi-row "INSERT ... VALUES (...), (...), ..." ?) - SELECT ... INSERT is not safe for statement-based replication, so will be logged as row in MIXED mode and give a warning in STATEMENT. But I wanted to hear you if you know any case and/or MariaDB version where a significant amount of auto-increment lock waits could be expected in parallel replication? And if so, maybe this fix should be pushed only to later MariaDB versions? - Kristian. Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: > revision-id: dbe5c20b755b87f67d87990c3baabc866667e41b (mariadb-10.4.30-2-gdbe5c20b755) > parent(s): a6114df595eeb7aeed4b050c9a3f4640c4320b5f > author: Kristian Nielsen > committer: Kristian Nielsen > timestamp: 2023-07-09 16:45:47 +0200 > message: > > MDEV-31482: Lock wait timeout with INSERT-SELECT, autoinc, and statement-based replication > > Remove the exception that InnoDB does not report auto-increment locks waits > to the parallel replication. > > There was an assumption that these waits could not cause conflicts with > in-order parallel replication and thus need not be reported. However, this > assumption is wrong and it is possible to get conflicts that lead to hangs > for the duration of --innodb-lock-wait-timeout. This can be seen with three > transactions: > > 1. T1 is waiting for T3 on an autoinc lock > 2. T2 is waiting for T1 to commit > 3. T3 is waiting on a normal row lock held by T2 > > Here, T3 needs to be deadlock killed on the wait by T1. > > Signed-off-by: Kristian Nielsen <knielsen(a)knielsen-hq.org> > > --- > mysql-test/suite/rpl/r/rpl_parallel_autoinc.result | 95 ++++++++++++++ > mysql-test/suite/rpl/t/rpl_parallel_autoinc.test | 140 +++++++++++++++++++++ > sql/sql_class.cc | 6 - > storage/innobase/lock/lock0lock.cc | 8 +- > 4 files changed, 236 insertions(+), 13 deletions(-) > > diff --git a/mysql-test/suite/rpl/r/rpl_parallel_autoinc.result b/mysql-test/suite/rpl/r/rpl_parallel_autoinc.result > new file mode 100644 > index 00000000000..c1829bafa1a > --- /dev/null > +++ b/mysql-test/suite/rpl/r/rpl_parallel_autoinc.result > @@ -0,0 +1,95 @@ > +include/master-slave.inc > +[connection master] > +MDEV-31482: Lock wait timeout with INSERT-SELECT, autoinc, and statement-based replication > +include/rpl_connect.inc [creating slave2] > +include/rpl_connect.inc [creating slave3] > +connection master; > +ALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB; > +CREATE TABLE t1 (a INT PRIMARY KEY AUTO_INCREMENT, b INT, c INT, INDEX (c)) ENGINE=InnoDB; > +INSERT INTO t1 (b,c) VALUES (0, 1), (0, 1), (0, 2), (0,3), (0, 5), (0, 7), (0, 8); > +CREATE TABLE t2 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB; > +INSERT INTO t2 VALUES (10,1), (20,2), (30,3), (40,4), (50,5); > +CREATE TABLE t3 (a VARCHAR(20) PRIMARY KEY, b INT) ENGINE=InnoDB; > +INSERT INTO t3 VALUES ('row for T1', 0), ('row for T2', 0), ('row for T3', 0); > +include/save_master_gtid.inc > +connection slave; > +include/sync_with_master_gtid.inc > +include/stop_slave.inc > +set @@global.slave_parallel_threads= 3; > +set @@global.slave_parallel_mode= OPTIMISTIC; > +set @@global.innodb_lock_wait_timeout= 20; > +connection master; > +BEGIN; > +UPDATE t3 SET b=b+1 where a="row for T1"; > +INSERT INTO t1(b, c) SELECT 1, t2.b FROM t2 WHERE a=10; > +Warnings: > +Note 1592 Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT. Statements writing to a table with an auto-increment column after selecting from another table are unsafe because the order in which rows are retrieved determines what (if any) rows will be written. This order cannot be predicted and may differ on master and the slave > +COMMIT; > +DELETE FROM t1 WHERE c >= 4 and c < 6; > +BEGIN; > +UPDATE t3 SET b=b+1 where a="row for T3"; > +INSERT INTO t1(b, c) SELECT 3, t2.b FROM t2 WHERE a >= 20 AND a <= 40; > +Warnings: > +Note 1592 Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT. Statements writing to a table with an auto-increment column after selecting from another table are unsafe because the order in which rows are retrieved determines what (if any) rows will be written. This order cannot be predicted and may differ on master and the slave > +COMMIT; > +include/save_master_gtid.inc > +connection slave1; > +BEGIN; > +SELECT * FROM t3 WHERE a="row for T1" FOR UPDATE; > +a b > +row for T1 0 > +connection slave2; > +BEGIN; > +SELECT * FROM t3 WHERE a="row for T3" FOR UPDATE; > +a b > +row for T3 0 > +connection slave3; > +BEGIN; > +DELETE FROM t2 WHERE a=30; > +connection slave; > +include/start_slave.inc > +connection slave2; > +ROLLBACK; > +connection slave1; > +ROLLBACK; > +connection slave3; > +ROLLBACK; > +connection slave; > +include/sync_with_master_gtid.inc > +SELECT * FROM t1 ORDER BY a; > +a b c > +1 0 1 > +2 0 1 > +3 0 2 > +4 0 3 > +6 0 7 > +7 0 8 > +8 1 1 > +9 3 2 > +10 3 3 > +11 3 4 > +SELECT * FROM t2 ORDER BY a; > +a b > +10 1 > +20 2 > +30 3 > +40 4 > +50 5 > +SELECT * FROM t3 ORDER BY a; > +a b > +row for T1 1 > +row for T2 0 > +row for T3 1 > +connection master; > +CALL mtr.add_suppression("Unsafe statement written to the binary log using statement format"); > +DROP TABLE t1, t2, t3; > +connection slave; > +include/stop_slave.inc > +SET @@global.slave_parallel_threads= 0; > +SET @@global.slave_parallel_mode= conservative; > +SET @@global.innodb_lock_wait_timeout= 50; > +include/start_slave.inc > +SELECT @@GLOBAL.innodb_autoinc_lock_mode; > +@@GLOBAL.innodb_autoinc_lock_mode > +1 > +include/rpl_end.inc > diff --git a/mysql-test/suite/rpl/t/rpl_parallel_autoinc.test b/mysql-test/suite/rpl/t/rpl_parallel_autoinc.test > new file mode 100644 > index 00000000000..0e96b4dfb80 > --- /dev/null > +++ b/mysql-test/suite/rpl/t/rpl_parallel_autoinc.test > @@ -0,0 +1,140 @@ > +--source include/have_binlog_format_statement.inc > +--source include/have_innodb.inc > +--source include/master-slave.inc > + > +--echo MDEV-31482: Lock wait timeout with INSERT-SELECT, autoinc, and statement-based replication > + > +# The scenario is transactions T1, T2, T3: > +# > +# T1 is waiting for T3 on an autoinc lock > +# T2 is waiting for T1 to commit > +# T3 is waiting on a normal row lock held by T2 > +# > +# This caused a hang until innodb_lock_wait_timeout, because autoinc > +# locks were not reported to the in-order parallel replication, so T3 > +# was not deadlock killed. > + > +--let $lock_wait_timeout=20 > + > +--let $rpl_connection_name= slave2 > +--let $rpl_server_number= 2 > +--source include/rpl_connect.inc > + > +--let $rpl_connection_name= slave3 > +--let $rpl_server_number= 2 > +--source include/rpl_connect.inc > + > +--connection master > +ALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB; > + > +# A table as destination for INSERT-SELECT > +CREATE TABLE t1 (a INT PRIMARY KEY AUTO_INCREMENT, b INT, c INT, INDEX (c)) ENGINE=InnoDB; > +INSERT INTO t1 (b,c) VALUES (0, 1), (0, 1), (0, 2), (0,3), (0, 5), (0, 7), (0, 8); > + > +# A table as source for INSERT-SELECT. > +CREATE TABLE t2 (a INT PRIMARY KEY, b INT) ENGINE=InnoDB; > +INSERT INTO t2 VALUES (10,1), (20,2), (30,3), (40,4), (50,5); > + > +# A table to help order slave worker threads to setup the desired scenario. > +CREATE TABLE t3 (a VARCHAR(20) PRIMARY KEY, b INT) ENGINE=InnoDB; > +INSERT INTO t3 VALUES ('row for T1', 0), ('row for T2', 0), ('row for T3', 0); > +--source include/save_master_gtid.inc > + > +--connection slave > +--source include/sync_with_master_gtid.inc > +--source include/stop_slave.inc > +--let $save_innodb_lock_wait_timeout= `SELECT @@global.innodb_lock_wait_timeout` > +--let $save_slave_parallel_threads= `SELECT @@global.slave_parallel_threads` > +--let $save_slave_parallel_mode= `SELECT @@global.slave_parallel_mode` > +set @@global.slave_parallel_threads= 3; > +set @@global.slave_parallel_mode= OPTIMISTIC; > +eval set @@global.innodb_lock_wait_timeout= $lock_wait_timeout; > + > +--connection master > +# Transaction T1. > +BEGIN; > +UPDATE t3 SET b=b+1 where a="row for T1"; > +INSERT INTO t1(b, c) SELECT 1, t2.b FROM t2 WHERE a=10; > +COMMIT; > + > +# Transaction T2. > +DELETE FROM t1 WHERE c >= 4 and c < 6; > + > +# Transaction T3. > +BEGIN; > +UPDATE t3 SET b=b+1 where a="row for T3"; > +INSERT INTO t1(b, c) SELECT 3, t2.b FROM t2 WHERE a >= 20 AND a <= 40; > +COMMIT; > + > +--source include/save_master_gtid.inc > + > +--connection slave1 > +# Temporarily block T1 to create the scheduling that triggers the bug. > +BEGIN; > +SELECT * FROM t3 WHERE a="row for T1" FOR UPDATE; > + > +--connection slave2 > +# Temporarily block T3 from starting (so T2 can reach commit). > +BEGIN; > +SELECT * FROM t3 WHERE a="row for T3" FOR UPDATE; > + > +--connection slave3 > +# This critical step blocks T3 after it has inserted its first row, > +# and thus taken the auto-increment lock, but before it has reached > +# the point where it gets a row lock wait on T2. Even though > +# auto-increment lock waits were not reported due to the bug, > +# transitive lock waits (T1 waits on autoinc of T3 which waits on row > +# on T2) _were_ reported as T1 waiting on T2, and thus a deadlock kill > +# happened and the bug was not triggered. > +BEGIN; > +DELETE FROM t2 WHERE a=30; > + > +--connection slave > +--source include/start_slave.inc > + > +# First let T2 complete until it is waiting for T1 to commit. > +--let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state='Waiting for prior transaction to commit' and command LIKE 'Slave_worker'; > +--source include/wait_condition.inc > + > +# Then let T3 reach the point where it has obtained the autoinc lock, > +# but it is not yet waiting for a row lock held by T2. > +--connection slave2 > +ROLLBACK; > +--let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state='Sending data' and info LIKE 'INSERT INTO t1(b, c) SELECT 3, t2.b%' and time_ms > 500 and command LIKE 'Slave_worker'; > +--source include/wait_condition.inc > + > +# Now let T1 continue, while T3 is holding the autoinc lock but before > +# it is waiting for T2. Wait a short while to give the hang a chance to > +# happen; T1 needs to get to request the autoinc lock before we let T3 > +# continue. (There's a small chance the sleep will be too small, which will > +# let the test occasionally pass on non-fixed server). > +--connection slave1 > +ROLLBACK; > +--sleep 0.5 > + > +# Now let T3 continue; the bug was that this lead to an undetected > +# deadlock that remained until innodb lock wait timeout. > +--connection slave3 > +ROLLBACK; > + > +--connection slave > +--let $slave_timeout= `SELECT $lock_wait_timeout/2` > +--source include/sync_with_master_gtid.inc > +--let $slave_timeout= > +SELECT * FROM t1 ORDER BY a; > +SELECT * FROM t2 ORDER BY a; > +SELECT * FROM t3 ORDER BY a; > + > +# Cleanup. > +--connection master > +CALL mtr.add_suppression("Unsafe statement written to the binary log using statement format"); > +DROP TABLE t1, t2, t3; > + > +--connection slave > +--source include/stop_slave.inc > +eval SET @@global.slave_parallel_threads= $save_slave_parallel_threads; > +eval SET @@global.slave_parallel_mode= $save_slave_parallel_mode; > +eval SET @@global.innodb_lock_wait_timeout= $save_innodb_lock_wait_timeout; > +--source include/start_slave.inc > +SELECT @@GLOBAL.innodb_autoinc_lock_mode; > +--source include/rpl_end.inc > diff --git a/sql/sql_class.cc b/sql/sql_class.cc > index 73bb654080a..8ed3f8a9c5e 100644 > --- a/sql/sql_class.cc > +++ b/sql/sql_class.cc > @@ -5119,12 +5119,6 @@ thd_need_wait_reports(const MYSQL_THD thd) > deadlock with the pre-determined commit order, we kill the later > transaction, and later re-try it, to resolve the deadlock. > > - This call need only receive reports about waits for locks that will remain > - until the holding transaction commits. InnoDB auto-increment locks, > - for example, are released earlier, and so need not be reported. (Such false > - positives are not harmful, but could lead to unnecessary kill and retry, so > - best avoided). > - > Returns 1 if the OTHER_THD will be killed to resolve deadlock, 0 if not. The > actual kill will happen later, asynchronously from another thread. The > caller does not need to take any actions on the return value if the > diff --git a/storage/innobase/lock/lock0lock.cc b/storage/innobase/lock/lock0lock.cc > index 26388ad95e2..b1cf2152cd6 100644 > --- a/storage/innobase/lock/lock0lock.cc > +++ b/storage/innobase/lock/lock0lock.cc > @@ -6872,13 +6872,7 @@ DeadlockChecker::search() > return m_start; > } > > - /* We do not need to report autoinc locks to the upper > - layer. These locks are released before commit, so they > - can not cause deadlocks with binlog-fixed commit > - order. */ > - if (m_report_waiters > - && (lock_get_type_low(lock) != LOCK_TABLE > - || lock_get_mode(lock) != LOCK_AUTO_INC)) { > + if (m_report_waiters) { > thd_rpl_deadlock_check(m_start->mysql_thd, > lock->trx->mysql_thd); > }

2 1

Proposal: mariadbd should automatically upgrade system tables and data directory on startup
by Otto Kekäläinen 03 Aug '23

03 Aug '23

Hi! I frequently see users running into various problems after upgrading MariaDB and starting the upgraded mariadbd server: - Server might refuse to start, asking users to manually run mariadb-upgrade - Server might start, but misbehave as on-disk data format for regular tables or the system tables has not updated - On systems where mariadb-upgrade runs automatically on MariaDB service startup there might be unnecessary delay if mariadb-upgrade is run multiple times - If running mariadb-upgrade fails or is slow, users might end up in a situation where they re-restart the server and have two mariadb-upgrade scripts running and blocking each other - In some cases user might attempt to run mariadb-upgrade but it fails it the MariaDB server was not running, forcing user to figure out how to manually start the server for upgrades - Users might end up starting server in a mode that does not accept connections with the intent to only run mariadb-upgrade, but then don't understand they need to restart the server to have it serve connections again - Users planning upgrades get confused because because reading about Fedora based distros the DBA is requires to manually run mariadb-upgrade themselves, and reading about Debian based distros DBAs are not required to do this as it is automatic, but they might end up trying to run it manually anyway because of prior experience with Fedora Some of the above issues can be remediated by running 'mariadb_upgrade --check-if-upgrade-is-needed' before actually running mariadb_upgrade in service scripts, but it does not work if server is offline (https://jira.mariadb.org/browse/MDEV-27636) Making users aware that their database is running but mariadb-upgrade has not run can be done via https://github.com/MariaDB/server/pull/2464. However this are just duct tape to the original problem. All of the above could be permanently solved if mariadbd simply did the upgrade by itself automatically. Can we please have this? Can we agree that this would be a desired feature in next MariaDB 11.x version, and if somebody implements this, it would be timely reviewed and merged? There is already design outlined at https://jira.mariadb.org/browse/MDEV-30499

8 19

Problem Upgrading Database
by jashley＠deadacated.com 25 Jul '23

25 Jul '23

Greetings, MariaDB 10.11.4 on Server 2022 Upgrading from 10.11.4 to 11.0.2 or 11.1.1 fails on Phase 5/10: Waiting for startup to complete. Is this a known problem? Thank you, mux

3 2

Re: 64ac6ed3e35: MDEV-31742 incorrect examined rows in case of stored function usage
by Sergei Golubchik 21 Jul '23

21 Jul '23

Hi, Oleksandr, On Jul 21, Oleksandr Byelkin wrote: > revision-id: 64ac6ed3e35 (mariadb-10.4.30-31-g64ac6ed3e35) > parent(s): 55a27bac80c > author: Oleksandr Byelkin > committer: Oleksandr Byelkin > timestamp: 2023-07-19 16:14:39 +0200 > message: > > MDEV-31742 incorrect examined rows in case of stored function usage > > The counter is global so we do not need add backup to it > if we do not zero it after taking the backup. > > diff --git a/sql/sql_class.cc b/sql/sql_class.cc > index c9e86b6142c..7911c78b6e0 100644 > --- a/sql/sql_class.cc > +++ b/sql/sql_class.cc > @@ -5568,7 +5568,6 @@ void THD::restore_sub_statement_state(Sub_statement_state *backup) > The following is added to the old values as we are interested in the > total complexity of the query > */ > - inc_examined_row_count(backup->examined_row_count); may be backup->examined_row_count should be simply removed? Sub_statement_state::examined_row_count, I mean. Not saved, not restored? There's also void THD::add_slow_query_state(Sub_statement_state *backup) { affected_rows+= backup->affected_rows; bytes_sent_old= backup->bytes_sent_old; m_examined_row_count+= backup->examined_row_count; it's not even using inc_examined_row_count(). Also examined_row_count is mentioned in comments in sql_class.cc, if you'll be removing it from Sub_statement_state, don't forget to fix comments to match. > cuted_fields+= backup->cuted_fields; > DBUG_VOID_RETURN; > } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 2

Re: 08403d85b00: MDEV-31677 Assertion failed upon online ALTER with binlog_row_image=NOBLOB
by Sergei Golubchik 21 Jul '23

21 Jul '23

Hi, Nikita, ok to push after making few small changes, see inline: On Jul 20, Nikita Malyavin wrote: > revision-id: 08403d85b00 (mariadb-11.0.1-147-g08403d85b00) > parent(s): a31b2a52119 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-07-19 19:08:51 +0400 > message: > > MDEV-31677 Assertion failed upon online ALTER with binlog_row_image=NOBLOB > > Make binlog_prepare_row_images accept image type as an argument. > > diff --git a/sql/sql_class.cc b/sql/sql_class.cc > index d4ad0efbf3a..bb2453af6a0 100644 > --- a/sql/sql_class.cc > +++ b/sql/sql_class.cc > @@ -7070,7 +7071,7 @@ int THD::binlog_update_row(TABLE* table, Event_log *bin_log, > not needed for binlogging. This is done according to the: > binlog-row-image option. the comment is outdated, should say This is done according to the row_image argument > */ > - binlog_prepare_row_images(table); > + binlog_prepare_row_images(table, row_image); > > size_t const before_maxlen= max_row_length(table, table->read_set, > before_record); > @@ -7194,7 +7195,7 @@ void THD::binlog_prepare_row_images(TABLE *table) > and the handler involved supports this. > */ > if (table->s->primary_key < MAX_KEY && > - (thd->variables.binlog_row_image < BINLOG_ROW_IMAGE_FULL) && > + row_image < BINLOG_ROW_IMAGE_FULL && > !ha_check_storage_engine_flag(table->s->db_type(), > HTON_NO_BINLOG_ROW_OPT)) > { please also repace thd->variables.binlog_row_image in the switch() below. doesn't make any practical difference, just for clarity Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Debugging MDEV-28776, the rare rpl.rpl_mark_optimize_tbl_ddl test failure
by Kristian Nielsen 21 Jul '23

21 Jul '23

Hi Brandon, The test failure in https://jira.mariadb.org/browse/MDEV-28776 (the one with 10 failed retries after deadlock, not the freebsd failures) seems like it could be something serious. But it has been very difficult to track down. I believe you have tried different things, and I also tried hard to reproduce it, so far without success. But it's still failing occasionally in actual buildbot runs. What do you think about pushing the below to the different branches? It makes the test case log InnoDB information about deadlocks encountered. This way, when it fails the next time in buildbot we can see 1) what kind of InnoDB deadlock is triggering the problem, if any; and 2) if it actually gets a deadlock inside InnoDB 10 times in a row, or if it is something else. - Kristian. ----------------------------------------------------------------------- Author: Kristian Nielsen <knielsen(a)knielsen-hq.org> Date: Sun Jul 9 15:18:03 2023 +0200 MDEV-28776: rpl.rpl_mark_optimize_tbl_ddl fails with timeout on sync_with_master This commit just extends the testcase to include some more information when the test failure occurs. The failure has so far only been reproducible in buildbot runs. Signed-off-by: Kristian Nielsen <knielsen(a)knielsen-hq.org> ----------------------------------------------------------------------- diff --git a/mysql-test/suite/rpl/t/rpl_mark_optimize_tbl_ddl.opt b/mysql-test/suite/rpl/t/rpl_mark_optimize_tbl_ddl.opt new file mode 100644 index 00000000000..3c53d5257b4 --- /dev/null +++ b/mysql-test/suite/rpl/t/rpl_mark_optimize_tbl_ddl.opt @@ -0,0 +1 @@ +--innodb-print-all-deadlocks=1

3 6

Re: 534031b77e0: MDEV-31400 Simple plugin dependency resolution
by Sergei Golubchik 20 Jul '23

20 Jul '23

Hi, Yuchen, Thanks. Conceptually all good, the approach is correct. See comments below - about the state and about user visible warnings, nothing major. On Jul 18, Yuchen Pei wrote: > revision-id: 534031b77e0 (mariadb-10.9.5-17-g534031b77e0) > parent(s): d8997f875e2 > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-07-03 17:54:56 +1000 > message: > > MDEV-31400 Simple plugin dependency resolution > > diff --git a/sql/sql_plugin.cc b/sql/sql_plugin.cc > index 5a077a934ac..15717352040 100644 > --- a/sql/sql_plugin.cc > +++ b/sql/sql_plugin.cc > @@ -1435,6 +1435,47 @@ void plugin_unlock_list(THD *thd, plugin_ref *list, size_t count) > DBUG_VOID_RETURN; > } > > +static int plugin_do_initialize(struct st_plugin_int *plugin, uint &state) > +{ > + DBUG_ENTER("plugin_do_initialize"); > + mysql_mutex_assert_not_owner(&LOCK_plugin); > + plugin_type_init init= plugin_type_initialize[plugin->plugin->type]; > + if (!init) > + init= (plugin_type_init) plugin->plugin->init; > + if (init) > + if (int ret= init(plugin)) > + { > + /* Plugin init failed but requested a retry if possible */ > + if (unlikely(ret== HA_ERR_RETRY_INIT)) > + sql_print_warning("Plugin '%s' registration as a %s failed.", > + plugin->name.str, plugin_type_names[plugin->plugin->type].str); > + else > + sql_print_error("Plugin '%s' registration as a %s failed.", > + plugin->name.str, plugin_type_names[plugin->plugin->type].str); > + DBUG_RETURN(ret); > + } > + state= PLUGIN_IS_READY; // plugin->init() succeeded > + > + if (plugin->plugin->status_vars) > + { > + /* > + historical ndb behavior caused MySQL plugins to specify > + status var names in full, with the plugin name prefix. > + this was never fixed in MySQL. > + MariaDB fixes that but supports MySQL style too. > + */ > + SHOW_VAR *show_vars= plugin->plugin->status_vars; > + SHOW_VAR tmp_array[2]= {{plugin->plugin->name, > + (char *) plugin->plugin->status_vars, SHOW_ARRAY}, > + {0, 0, SHOW_UNDEF}}; > + if (strncasecmp(show_vars->name, plugin->name.str, plugin->name.length)) > + show_vars= tmp_array; > + > + if (add_status_vars(show_vars)) > + DBUG_RETURN(1); > + } > + DBUG_RETURN(0); > +} > > static int plugin_initialize(MEM_ROOT *tmp_root, struct st_plugin_int *plugin, > int *argc, char **argv, bool options_only) > @@ -1595,7 +1594,7 @@ int plugin_init(int *argc, char **argv, int flags) > size_t i; > struct st_maria_plugin **builtins; > struct st_maria_plugin *plugin; > - struct st_plugin_int tmp, *plugin_ptr, **reap; > + struct st_plugin_int tmp, *plugin_ptr, **reap, **retry_end, **retry_start; > MEM_ROOT tmp_root; > bool reaped_mandatory_plugin= false; > bool mandatory= true; > @@ -1737,11 +1736,16 @@ int plugin_init(int *argc, char **argv, int flags) > */ > > mysql_mutex_lock(&LOCK_plugin); > + /* List of plugins to reap */ > reap= (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); > *(reap++)= NULL; > + /* List of plugins to retry */ > + retry_start= retry_end= > + (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); > > for(;;) > { > + int error; > for (i=0; i < MYSQL_MAX_PLUGIN_TYPE_NUM; i++) > { > HASH *hash= plugin_hash + plugin_type_initialization_order[i]; > @@ -1753,16 +1757,46 @@ int plugin_init(int *argc, char **argv, int flags) > bool plugin_table_engine= lex_string_eq(&plugin_table_engine_name, > &plugin_ptr->name); > bool opts_only= flags & PLUGIN_INIT_SKIP_INITIALIZATION && > - (flags & PLUGIN_INIT_SKIP_PLUGIN_TABLE || > - !plugin_table_engine); > - if (plugin_initialize(&tmp_root, plugin_ptr, argc, argv, opts_only)) > + (flags & PLUGIN_INIT_SKIP_PLUGIN_TABLE || !plugin_table_engine); > + error= plugin_initialize(&tmp_root, plugin_ptr, argc, argv, > + opts_only); > + if (error) > { > plugin_ptr->state= PLUGIN_IS_DYING; > - *(reap++)= plugin_ptr; > + /* The plugin wants a retry of the initialisation, > + possibly due to dependency on other plugins */ > + if (unlikely(error == HA_ERR_RETRY_INIT)) > + *(retry_end++)= plugin_ptr; > + else > + *(reap++)= plugin_ptr; > } > } > } > } > + /* Retry plugins that asked for it */ > + while (retry_start < retry_end) > + { > + st_plugin_int **to_re_retry, **retrying; > + for (to_re_retry= retrying= retry_start; retrying < retry_end; retrying++) > + { > + plugin_ptr= *retrying; > + uint state= plugin_ptr->state; > + mysql_mutex_unlock(&LOCK_plugin); you might want to add here sql_print_information("plugin XXX registration is retried") otherwise it would be confusing to see "plugin XXX registration failed" and then see the plugin perferctly initialized and usable. Alternatively (preferable, even) don't print "plugin registration failed" above, print it when adding a plugin to the reap list. So that the user won't see fail/retry loop at all, this is the inner working of the plugin subsystem, no user serviceable parts inside. > + error= plugin_do_initialize(plugin_ptr, state); > + mysql_mutex_lock(&LOCK_plugin); > + plugin_ptr->state= state; why do you need to manipulate (save/restore) the state here? Is that even correct? You set the state to PLUGIN_IS_DYING earlier and plugin_do_initialize() an change it to PLUGIN_IS_READY on success. You wouldn't want to change PLUGIN_IS_READY to PLUGIN_IS_DYING, would you? If this is the case, then you don't need to save the state at all, instead you need something like if (error) plugin_ptr->state= PLUGIN_IS_DYING; > + if (error == HA_ERR_RETRY_INIT) > + *(to_re_retry++)= plugin_ptr; > + else if (error) > + *(reap++)= plugin_ptr; > + } > + /* If the retry list has not changed, i.e. if all retry attempts > + result in another retry request, empty the retry list */ > + if (to_re_retry == retry_end) > + while (to_re_retry > retry_start) > + *(reap++)= *(--to_re_retry); > + retry_end= to_re_retry; > + } > > /* load and init plugins from the plugin table (unless done already) */ > if (flags & PLUGIN_INIT_SKIP_PLUGIN_TABLE) > @@ -1775,7 +1809,8 @@ int plugin_init(int *argc, char **argv, int flags) > } > > /* > - Check if any plugins have to be reaped > + Merge the retry list to the reap list, then reap the failed > + plugins. Note that during the merge we reverse the order in retry seems like this comment is obsolete > */ > while ((plugin_ptr= *(--reap))) > { > @@ -1788,6 +1823,7 @@ int plugin_init(int *argc, char **argv, int flags) > } > > mysql_mutex_unlock(&LOCK_plugin); > + my_afree(retry_start); > my_afree(reap); > if (reaped_mandatory_plugin && !opt_help) > goto err; > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: Client-server protocol improvement proposal: Connection redirection
by Andrey Sidorov 19 Jul '23

19 Jul '23

Hi Otto, thanks for bringing this to my attention. Happy to add redirect support to mysql2, looks like its relatively straightforward and sits on top of the existing protocol Andrey On Wed, 19 Jul 2023 at 15:44, Otto Kekäläinen <otto(a)kekalainen.net> wrote: > Hi Andrey! > > I am reaching out to you because you maintain > https://github.com/sidorares/node-mysql2 > > Have you heard about this proposal on the MariaDB developers list? > > It suggests adding a new feature to the MariaDB/MySQL protocol where > the server would be able to tell clients "No accepting connections - > connect to this server instead". The implementation would be work by > sending a new error type ERR_REDIRECT along with a client the > IP/hostname of a new server, similar to how HTTP 301 redirect works- > > Full thread visible at > > https://lists.mariadb.org/hyperkitty/list/developers@lists.mariadb.org/thre… > > If you want to participate in discussion, just reply to this message > and change recipient to developers(a)lists.mariadb.org. > > Thanks! > > > ---------- Forwarded message --------- > From: Lenski, Daniel via developers <developers(a)lists.mariadb.org> > Date: Wed, 28 Jun 2023 at 15:24 > Subject: [MariaDB developers] Client-server protocol improvement > proposal: Connection redirection > To: developers(a)lists.mariadb.org <developers(a)lists.mariadb.org> > > > Hi! > > I would like to propose a new feature in the MariaDB client-server > protocol: application-layer redirection of client connections. > > We want the MariaDB server to be able to tell clients connecting to > it, “Sorry, this server is unavailable. Connect to an alternate server > instead.” This mechanism is inspired by HTTP 302 (“temporary > redirect”) mechanism familiar to all developers of web applications, > and is intended to have similar semantics and security properties, > since these have now been widely deployed and tested for decades. > > I have submitted a minimal but viable implementation of this at: > https://github.com/MariaDB/server/pull/2681, server-side implementation > https://github.com/mariadb-corporation/mariadb-connector-c/pull/226, > MariaDB Connector/C implementation > > > I am seeking advice and opinions on these, and will proceed to > finalize this in the following 1-2 months. > > In its current form, this implementation allows the MariaDB server > administrator to set two variables: > > SET GLOBAL SERVER_REDIRECT_MODE={ON,OFF} (the default is OFF) > SET GLOBAL SERVER_REDIRECT_TARGET='my-new-server.example.com' (or > 192.168.0.123:3307, or new-server.com:3308, etc) > > When SERVER_REDIRECT_MODE is set to on, the server will stop accepting > new TCP-based connections (unless they are to the extra port, which is > intended for emergency administrative use), and it will instead > respond to all clients with an error packet: > > $ mariadb --host my-server.example.com > ERROR 4189 (HY000): |Server is redirecting clients to > 'my-new-server.example.com:3307'|my-new-server.example.com:3307 > > When an appropriately-updated client receives this error packet (error > code 4196, message formatted as |Human readable message|<value of > SERVER_REDIRECT_TARGET>), it will parse the redirection target out > from the error message and attempt a new connection: > > $ updated-mariadb --host my-server.example.com -e 'select @@hostname' > Got server redirect to 'my-new-server.example.com' (port 3307) > +---------------------------+ > | @@hostname | > +---------------------------+ > | my-new-server.example.com | > +---------------------------+ > > The feature is gracefully backwards-compatible. Old clients that do > not support redirection will fail and show the message in a form from > which humans can deduce what happened. New clients that do support it > will follow it to the new server. > > [...] >

1 0

Re: 845fbe9ce01: MDEV-30820 slow log Rows_examined out of range
by Sergei Golubchik 18 Jul '23

18 Jul '23

Hi, Sanja, The patch is ok, but the test case is rather weird. What does it do? How long does it run? On Jul 17, Oleksandr Byelkin wrote: > revision-id: 845fbe9ce01 (mariadb-10.4.30-30-g845fbe9ce01) > parent(s): 02cd3675c4d > author: Oleksandr Byelkin > committer: Oleksandr Byelkin > timestamp: 2023-07-17 10:22:54 +0200 > message: > > MDEV-30820 slow log Rows_examined out of range > > Fix row counters to be able to get any possible value. > > diff --git a/mysql-test/main/log_slow.test b/mysql-test/main/log_slow.test > --- a/mysql-test/main/log_slow.test > +++ b/mysql-test/main/log_slow.test > @@ -119,3 +119,98 @@ drop table t; > --echo # > --echo # End of 10.3 tests > --echo # > + > + > +--echo # > +--echo # MDEV-30820: slow log Rows_examined out of range > +--echo # > + > +set @log_output.save= @@global.log_output, @slow_log.save= @@global.slow_query_log; > +set global log_output= 'TABLE', slow_query_log= ON; > +set long_query_time= 0.000001; > +create database db; > +use db; > + > +--delimiter // > +CREATE OR REPLACE FUNCTION `get_id`(`INPUT_STRING` VARCHAR(50), `DECRYPT_KEY` CHAR(20)) RETURNS int(11) > +BEGIN > + DECLARE > + REQUEST_ID INT ; > + SET > + REQUEST_ID = > + IF( > + ( > + LOCATE('#$#$#$ LHP_ID_', INPUT_STRING) > + ) > 0 AND > + ( > + INPUT_STRING REGEXP DECRYPT_KEY > + )>0, > + ( > + SPLIT_STRING( > + SUBSTRING_INDEX( > + SUBSTRING_INDEX(INPUT_STRING, '#$#$#$ LHP_ID_', -1), > + '#$#$#$', > + 1 > + ), > + '-', > + 2 > + ) > + ), > + NULL > + ) ; > + SET > + REQUEST_ID=IF( > + ISNULL(REQUEST_ID)OR(REQUEST_ID='')OR(REQUEST_ID=0), > + NULL, > + REQUEST_ID > + > + ); > + RETURN(REQUEST_ID) ; > +END > +// > + > +CREATE OR REPLACE FUNCTION `SPLIT_STRING`(`str` VARCHAR(255), `delim` VARCHAR(12), `pos` INT) RETURNS varchar(255) CHARSET utf8mb4 > +RETURN REPLACE( > + SUBSTRING( > + SUBSTRING_INDEX(str , delim , pos) , > + CHAR_LENGTH( > + SUBSTRING_INDEX(str , delim , pos - 1) > + ) + 1 > + ) , > + delim , > + '' > +) > +// > + > +CREATE TABLE `tab1` ( > +`ID` int(11) NOT NULL AUTO_INCREMENT, > +`NAME_F` varchar(50) DEFAULT NULL, > + PRIMARY KEY (`ID`) > +) DEFAULT CHARSET=utf8mb4 // > + > + CREATE TABLE `tab2` ( > + `ID` int(11) NOT NULL AUTO_INCREMENT, > + `TAB1_ID` int(11) DEFAULT NULL, > + PRIMARY KEY (`id`) > +) DEFAULT CHARSET=utf8mb4 // > + > +for i in 1..100 do insert into tab1 values (null,'qwerty'); end for ; // > +for i in 1..1000 do insert into tab2 values (null,round(rand()*10000)); end for ; // > + > +--delimiter ; > + > +SELECT > + get_id(CAST(aes_decrypt(tab1.NAME_F,'V41iNM0n4') AS char),'KM_ID_PL') as get_string, > + (CASE WHEN (SELECT ID FROM tab2 where tab2.TAB1_ID = tab1.ID LIMIT 1) IS NULL THEN 0 ELSE 1 END) AS IS_ATT > +FROM > + tab1 > +ORDER BY 2 DESC > +LIMIT 2; > + > +set global log_output= @log_output.save, slow_query_log= @slow_log.save; > +drop database db; > + > +--echo # > +--echo # End of 10.4 tests > +--echo # Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 5

Re: 5102d86beae: MDEV-30164 System variable for default collations
by Sergei Golubchik 17 Jul '23

17 Jul '23

Hi, Alexander, Looks good, thanks. ok to push. a couple of comments below - but really, on remnants of the previous version of the commit. On Jul 17, Alexander Barkov wrote: > revision-id: 5102d86beae (mariadb-10.11.2-47-g5102d86beae) > parent(s): 3430767e006 > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-07-10 13:17:45 +0400 > message: > > MDEV-30164 System variable for default collations > > This patch adds a way to override default collations > (or "character set collations") for desired character sets. > > diff --git a/mysql-test/main/mysqld--help.result b/mysql-test/main/mysqld--help.result > index 77df8536fa2..5c361e3da88 100644 > --- a/mysql-test/main/mysqld--help.result > +++ b/mysql-test/main/mysqld--help.result > @@ -144,6 +144,8 @@ The following specify which files/extra groups are read (specified before remain > Don't ignore client side character set value sent during > handshake. > (Defaults to on; use --skip-character-set-client-handshake to disable.) > + --character-set-collations=name > + Set default collations for character sets. shouldn't it say "overrides"? > --character-set-filesystem=name > Set the filesystem character set. > -C, --character-set-server=name > diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc > index 607877dfb83..d35869e7e39 100644 > --- a/sql/sql_parse.cc > +++ b/sql/sql_parse.cc > @@ -6061,6 +6061,32 @@ mysql_execute_command(THD *thd, bool is_called_from_prepared_stmt) > } > thd->reset_kill_query(); > } > + > + /* > + Binary logging is now done. Unset the "used" flags to avoid > + flags leaking to the next event (and to the COMMIT statement > + in the end of the current event). > + > + Example: > + > + Suppose a non-default collation (in @@character_set_collations) > + was used during the statement, the mysqlbinlog output for > + the current statement will contain a sequence like this: > + > + SET character_set_collations='utf8mb3=utf8mb3_bin'; > + INSERT INTO t1 VALUES (_utf8mb3'test'); > + COMMIT; > + > + The statment (INSERT in this example) is already in binlog at this point, and the > + and the "SET character_set_collations" is written inside a > + Q_CHARACTER_SET_COLLATIONS chunk in its log entry header. > + The flag CHARACTER_SET_COLLATIONS_USED is not needed any more. > + The COMMIT can be printed without "SET character_set_collations". > + > + The same logic applies to the other _USED flags. That's a bit too verbose in my opinion. The example was good in the previous version of the patch, but now the first paragraph of the comment pretty much explains it all. but do it as you prefer. > + */ > + thd->used= 0; > + > if (unlikely(thd->is_error()) || > (thd->variables.option_bits & OPTION_MASTER_SQL_ERROR)) > { Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 8e9c68b49dd: MDEV-25237 Assertion `global_system_variables.session_track_system_variables' failed in Session_sysvars_tracker::init | SIGSEGV's in __strlen_avx2 | UBSAN: runtime error: null pointer passed as argument 1, which is declared to never be null in my_strdup
by Sergei Golubchik 15 Jul '23

15 Jul '23

Hi, Sanja, ok to push. but see comments below: On Jul 15, Oleksandr Byelkin wrote: > revision-id: 8e9c68b49dd (mariadb-10.4.30-9-g8e9c68b49dd) > parent(s): 423c28f0aa4 > author: Oleksandr Byelkin > committer: Oleksandr Byelkin > timestamp: 2023-07-10 13:40:46 +0200 > message: > > MDEV-25237 Assertion `global_system_variables.session_track_system_variables' failed in Session_sysvars_tracker::init | SIGSEGV's in __strlen_avx2 | UBSAN: runtime error: null pointer passed as argument 1, which is declared to never be null in my_strdup again, if you see a bug with a very long summary, particularly if it basically makes no sense to a user - in that case, please, don't hesitate to make it shorter and more meaningful. For example "crash after setting session_track_system_variables to an invalid value" > Fix of typo in checking variable list corectness. > > Fix of error handling in case of variable list parse error > > diff --git a/sql/session_tracker.cc b/sql/session_tracker.cc > index 2f6c5e29bf2..3e1558f6759 100644 > --- a/sql/session_tracker.cc > +++ b/sql/session_tracker.cc > @@ -221,7 +221,7 @@ bool sysvartrack_validate_value(THD *thd, const char *str, size_t len) > /* Remove leading/trailing whitespace. */ > trim_whitespace(system_charset_info, &var); > > - if (!strcmp(var.str, "*") && !find_sys_var(thd, var.str, var.length)) > + if (strcmp(var.str, "*") && !find_sys_var(thd, var.str, var.length)) > return true; > > if (lasts) > @@ -331,9 +331,10 @@ void Session_sysvars_tracker::init(THD *thd) > mysql_mutex_assert_owner(&LOCK_global_system_variables); > DBUG_ASSERT(thd->variables.session_track_system_variables == > global_system_variables.session_track_system_variables); > - DBUG_ASSERT(global_system_variables.session_track_system_variables); > thd->variables.session_track_system_variables= > - my_strdup(global_system_variables.session_track_system_variables, > + my_strdup((global_system_variables.session_track_system_variables? > + global_system_variables.session_track_system_variables: > + ""), this is such a common pattern that we have a helper for that, use my_strdup(safe_str(global_system_variables.session_track_system_variables)) > MYF(MY_WME | MY_THREAD_SPECIFIC)); > } > > @@ -572,6 +573,12 @@ bool sysvartrack_global_update(THD *thd, char *str, size_t len) > { > LEX_STRING tmp= { str, len }; > Session_sysvars_tracker::vars_list dummy; > + DBUG_EXECUTE_IF("dbug_session_tracker_parse_error", > + { > + my_error(ER_OUTOFMEMORY, MYF(0), 1); > + return true; > + }); > + > if (!dummy.parse_var_list(thd, tmp, false, system_charset_info)) > { > dummy.construct_var_list(str, len + 1); > diff --git a/sql/sys_vars.inl b/sql/sys_vars.inl > index 84d1cd6b331..3e282de439a 100644 > --- a/sql/sys_vars.inl > +++ b/sql/sys_vars.inl > @@ -620,7 +620,11 @@ class Sys_var_sesvartrack: public Sys_var_charptr_base > { > if (sysvartrack_global_update(thd, new_val, > var->save_result.string_value.length)) > + { > + if (new_val) > + my_free(new_val); it's ok, as you like. But technically you don't need an `if` here, my_free() just as free() works with NULL pointers fine. > new_val= 0; > + } > } > global_update_finish(new_val); > return (new_val == 0 && var->save_result.string_value.str != 0); > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: 8d51c6d234b: MDEV-30164 System variable for default collations
by Sergei Golubchik 10 Jul '23

10 Jul '23

Hi, Alexander, Please, see comments below. Sorry, it took a while. On Jun 23, Alexander Barkov wrote: > revision-id: 8d51c6d234b (mariadb-10.11.2-13-g8d51c6d234b) > parent(s): ceb0e7f944b > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-03-21 10:07:57 +0400 > message: > > MDEV-30164 System variable for default collations > > diff --git a/mysql-test/main/ctype_collate_implicit.test b/mysql-test/main/ctype_collate_implicit.test > --- /dev/null > +++ b/mysql-test/main/ctype_collate_implicit.test > @@ -0,0 +1,209 @@ > +--source include/have_utf8.inc > +--source include/have_utf8mb4.inc > + > +--echo # > +--echo # MDEV-30164 System variable for default collations > +--echo # > + > +SET @@character_set_collations= ' utf8mb3 = utf8mb3_bin , LATIN1 = LATIN1_BIN '; also, please, test that parsing of commas is relaxed too. Like set @@character_set_collations= ',,,utf8mb3 = utf8mb3_bin,,latin1 = latin1_bin,,,'; it makes it easier for the user to edit the value. > +SELECT @@character_set_collations; > +SET @@character_set_collations=''; > +SELECT @@character_set_collations; > + > +SET @@character_set_collations='utf8mb3=utf8mb3_bin'; > +SELECT @@character_set_collations; > +SET @@character_set_collations=''; > + > +--error ER_COLLATION_CHARSET_MISMATCH > +SET @@character_set_collations='utf8mb3=utf8mb4_general_ci'; > +SELECT @@character_set_collations; > + > +--error ER_COLLATION_CHARSET_MISMATCH > +SET @@character_set_collations='utf8mb4=utf8mb3_general_ci'; > +SELECT @@character_set_collations; > + > +SET @@character_set_collations='utf8mb3=utf8mb3_general_ci'; > +SELECT @@character_set_collations; > + > +SET @@character_set_collations='utf8mb4=utf8mb4_general_ci,latin1=latin1_bin'; > +SELECT @@character_set_collations; > + > +--error ER_COLLATION_CHARSET_MISMATCH > +SET @@character_set_collations='utf8mb4=uca1400_ai_ci,latin1=uca1400_ai_ci'; > + > +# All or nothing is set. "Nothing" in this case because of the error on latin1. > +# The "uca1400_ai_ci FOR utf8mb4" part was ignored. > +SELECT @@character_set_collations; > +SELECT @@character_set_collations RLIKE 'utf8mb4=utf8mb4_general_ci' AS expect_true; > + > + > +SET @@character_set_collations='utf8mb4=uca1400_ai_ci'; > +SELECT @@character_set_collations; > + > +SET NAMES utf8mb4; > +SELECT @@collation_connection; > + > +# We have to disable --view-protocol for the following statement. > +# 'mtr --view-protocol' creates a separate connection for these statements: > +# CREATE VIEW mysqltest_tmp_sp AS ...; > +# DROP VIEW mysqltest_tmp_sp; > +# The current @@character_set_collations does not affect this connection. > +# So --view-protocol would return the hard-coded character set collation here, > +# instead of utf8mb4_uca1400_ai_ci > + > +--disable_view_protocol > +SELECT collation('literal'); > +--enable_view_protocol > +EXECUTE IMMEDIATE 'SELECT COLLATION(?)' USING 'literal'; > + > +CREATE VIEW v1 AS SELECT 'literal', collation('literal') as cl; > +SHOW CREATE VIEW v1; > +SELECT * FROM v1; > +DROP VIEW v1; > + > + > +# Override @@collation_connection to utf8mb4_general_ci. > +# Make sure that CREATE statements does not use @@collation_connection. > +# to detect implicit collations. > +# Implicit collations are detected using @@character_set_collations! > + > +SET NAMES utf8mb4 COLLATE utf8mb4_general_ci; > + > +CREATE TABLE t1 (a TEXT CHARACTER SET utf8mb4); > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > + > +CREATE TABLE t1 (a TEXT CHARACTER SET utf8mb4 COLLATE DEFAULT); > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > + > +CREATE TABLE t1 (a TEXT COLLATE DEFAULT) CHARACTER SET utf8mb4; > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > + > +CREATE TABLE t1 (a TEXT) CHARACTER SET utf8mb4; > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > + > +CREATE DATABASE db1 CHARACTER SET utf8mb4; > +SHOW CREATE DATABASE db1; > +DROP DATABASE db1; > + > + > +# Test how @@character_set_collations affects various expressions > +# with implicit collations. > + > + > +let query=SELECT > + @@collation_connection AS conn, > + COLLATION('a') AS lit, > + COLLATION(CONCAT(1)) AS num, > + COLLATION(CAST(123 AS CHAR)) AS casti, > + COLLATION(_utf8mb4'a') AS litu, > + COLLATION(_utf8mb4 0x62) AS lituh, > + COLLATION(_utf8mb4 X'63') AS lituhs, > + COLLATION(CAST(123 AS CHAR CHARACTER SET utf8mb4)) AS castic, > + COLLATION(CHAR(0x61 USING utf8mb4)) AS chr, > + COLLATION(CONVERT('a' USING utf8mb4)) AS conv; > + > +# The below SET NAMES sets @@collation_connection to utf8mb4_general_ci. > +# But @@character_set_collations still contains utf8mb4=uca1400_ai_ci. > + > +SET NAMES utf8mb4 COLLATE utf8mb4_general_ci; > + > +# Columns expected to print utf8mb4_general_ci > +# because they use @@collation_connection: > +# - String literals without introducers > +# - Automatic number-to-string conversions > +# - CAST(AS CHAR) - without USING > +# > +# Columns expected to print utf8mb4_uca1400_ai_ci > +# because they use the current session default collation > +# for the character set (as specified in @@collation_connection) > +# - String literals with introducers > +# - CAST(AS CHAR USING cs) > +# - CHAR() > +# - CONVERT() > + > +--vertical_results > +--eval $query; > +--horizontal_results > + > +# This sets collation_connection to utf8mb4_uca1400_ai_ci > +# according to @@character_set_collations. > +SET NAMES utf8mb4; > + > +# Now all columns are expected to print utf8mb4_uca1400_ai_ci: > +# - Some columns because @@collation_connection says so > +# - Some columns because @@character_set_collations says so. > + > +--vertical_results > +--eval $query; > +--horizontal_results > + > + > +# > +# INFORMATION_SCHEMA > +# > + > +SET character_set_collations='latin1=latin1_bin,utf8mb4=uca1400_ai_ci'; > +SHOW CHARACTER SET LIKE 'latin1'; > +SELECT * FROM INFORMATION_SCHEMA.CHARACTER_SETS > +WHERE CHARACTER_SET_NAME='latin1'; > + > +SHOW COLLATION LIKE 'latin1%'; > +SELECT COLLATION_NAME, IS_DEFAULT > +FROM INFORMATION_SCHEMA.COLLATIONS > +WHERE CHARACTER_SET_NAME LIKE 'latin1%'; > +SELECT COLLATION_NAME, FULL_COLLATION_NAME, IS_DEFAULT > +FROM INFORMATION_SCHEMA.COLLATION_CHARACTER_SET_APPLICABILITY > +WHERE COLLATION_NAME LIKE 'latin1%'; > + > +SHOW CHARACTER SET LIKE 'utf8mb4'; > +SELECT * FROM INFORMATION_SCHEMA.CHARACTER_SETS > +WHERE CHARACTER_SET_NAME='utf8mb4'; > + > +SHOW COLLATION LIKE '%uca1400_ai_ci%'; > +SELECT COLLATION_NAME, IS_DEFAULT > +FROM INFORMATION_SCHEMA.COLLATIONS > +WHERE COLLATION_NAME LIKE '%uca1400_ai_ci%'; > +SELECT COLLATION_NAME, FULL_COLLATION_NAME, IS_DEFAULT > +FROM INFORMATION_SCHEMA.COLLATION_CHARACTER_SET_APPLICABILITY > +WHERE COLLATION_NAME LIKE '%uca1400_ai_ci%'; > + > +# > +# Prepared statements: reprepare on @@character_set_collations change. > +# > + > +SET @@character_set_collations=''; > +PREPARE stmt FROM 'SELECT ' > + 'COLLATION(CAST("x" AS CHAR CHARACTER SET utf8mb3)) AS a, ' > + 'COLLATION(_utf8mb3"x") AS b'; > +EXECUTE stmt; > +SET @@character_set_collations='utf8mb3=utf8mb3_bin'; > +EXECUTE stmt; > + > +SET @@character_set_collations='utf8mb3=utf8mb3_bin'; > +PREPARE stmt FROM 'SELECT ' > + 'COLLATION(CAST("x" AS CHAR CHARACTER SET utf8mb3)) AS a, ' > + 'COLLATION(_utf8mb3"x") AS b'; > +EXECUTE stmt; > +SET @@character_set_collations=DEFAULT; > +EXECUTE stmt; > + > +SET NAMES utf8mb3; > +SET @@character_set_collations=''; > +PREPARE stmt FROM 'CREATE TABLE t1 ' > + '(a TEXT CHARACTER SET utf8mb3 COLLATE DEFAULT COLLATE utf8mb3_general_ci)'; > +EXECUTE stmt; > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > + > +SET @@character_set_collations='utf8mb3=utf8mb3_bin'; > +--error ER_CONFLICTING_DECLARATIONS > +EXECUTE stmt; > + > +SET @@character_set_collations=''; > +EXECUTE stmt; > +SHOW CREATE TABLE t1; > +DROP TABLE t1; > diff --git a/mysql-test/suite/sys_vars/r/sysvars_server_notembedded.result b/mysql-test/suite/sys_vars/r/sysvars_server_notembedded.result > --- a/mysql-test/suite/sys_vars/r/sysvars_server_notembedded.result > +++ b/mysql-test/suite/sys_vars/r/sysvars_server_notembedded.result > @@ -502,6 +502,16 @@ NUMERIC_BLOCK_SIZE NULL > ENUM_VALUE_LIST NULL > READ_ONLY NO > COMMAND_LINE_ARGUMENT NULL > +VARIABLE_NAME CHARACTER_SET_COLLATIONS > +VARIABLE_SCOPE SESSION > +VARIABLE_TYPE VARCHAR > +VARIABLE_COMMENT Default collations for character sets Better to emphasize somehow that it's not a complete list of default collations, but adjustments to the compiled-in list of default collations. If this variable is empty, it doesn't mean that character sets have no default collations. I'm not sure how to put this succintly in the help text though :) > +NUMERIC_MIN_VALUE NULL > +NUMERIC_MAX_VALUE NULL > +NUMERIC_BLOCK_SIZE NULL > +ENUM_VALUE_LIST NULL > +READ_ONLY NO > +COMMAND_LINE_ARGUMENT NULL > VARIABLE_NAME CHARACTER_SET_CONNECTION > VARIABLE_SCOPE SESSION > VARIABLE_TYPE ENUM > diff --git a/sql/lex_charset.cc b/sql/lex_charset.cc > --- a/sql/lex_charset.cc > +++ b/sql/lex_charset.cc > @@ -447,6 +451,17 @@ Lex_exact_charset_opt_extended_collate::find_default_collation() const > } > > > +CHARSET_INFO * > +Lex_exact_charset_opt_extended_collate:: > + find_mapped_default_collation(Charset_collation_map_st::Used *used, > + const Charset_collation_map_st &map) const > +{ > + CHARSET_INFO *cs= find_compiled_default_collation(); > + if (!cs) > + return nullptr; > + return map.get_collation_for_charset(used, cs); This seems a bit redundant. find_compiled_default_collation() will do get_charset_by_csname(m_ci, MY_CS_PRIMARY) and only then you'll be able to search, because you map default compiled collation id to use specified collation id. but conceptually you need to map a charset to a collation, not a collation to a collation. If you'd map, say, cs->cs_name.str (just the pointer, not the string value) or, say, cs->cset, then you'd be mapping a charset to a collation. and here you'd be able to do just return map.get_collation_for_charset(used, m_ci) with find_compiled_default_collation() being done inside get_collation_for_charset() and only when Elem_st wasn't found > +} > + > /* > Resolve an empty or a contextually typed collation according to the > upper level default character set (and optionally a collation), e.g.: > diff --git a/sql/lex_charset.h b/sql/lex_charset.h > --- a/sql/lex_charset.h > +++ b/sql/lex_charset.h > @@ -544,6 +561,21 @@ struct Lex_exact_charset_extended_collation_attrs_st > { > return m_ci; > } > + CHARSET_INFO *charset_info(Charset_collation_map_st::Used *used, > + const Charset_collation_map_st &map) const > + { > + switch (m_type) > + { > + case TYPE_CHARACTER_SET: > + return map.get_collation_for_charset(used, m_ci); Why do you need this special charset_info(used,map) that differs from charset_info() when m_type==TYPE_CHARACTER_SET, and is used in set_charset_collation_attrs() only? > + case TYPE_EMPTY: > + case TYPE_CHARACTER_SET_COLLATE_EXACT: > + case TYPE_COLLATE_CONTEXTUALLY_TYPED: > + case TYPE_COLLATE_EXACT: > + break; > + } > + return m_ci; > + } > Type type() const > { > return m_type; > diff --git a/sql/log_event_server.cc b/sql/log_event_server.cc > --- a/sql/log_event_server.cc > +++ b/sql/log_event_server.cc > @@ -1989,6 +1997,16 @@ int Query_log_event::do_apply_event(rpl_group_info *rgi, > thd->variables.sql_mode= > (sql_mode_t) ((thd->variables.sql_mode & MODE_NO_DIR_IN_CREATE) | > (sql_mode & ~(sql_mode_t) MODE_NO_DIR_IN_CREATE)); > + > + size_t cslen= thd->variables.character_set_collations.from_binary( > + character_set_collations.str, > + character_set_collations.length); > + if (cslen != character_set_collations.length) > + { > + thd->variables.character_set_collations.init(); > + goto compare_errors; // QQ: report an error here? is it corrupted event or unknown collation? > + } > + > if (charset_inited) > { > rpl_sql_thread_info *sql_info= thd->system_thread_info.rpl_sql_info; > diff --git a/sql/sql_parse.cc b/sql/sql_parse.cc > --- a/sql/sql_parse.cc > +++ b/sql/sql_parse.cc > @@ -6049,6 +6049,27 @@ mysql_execute_command(THD *thd, bool is_called_from_prepared_stmt) > } > thd->reset_kill_query(); > } > + > + /* > + If a non-default collation (in @@character_set_collations) > + was used during the statement, the mysqlbinlog output for > + the current statement will contain a sequence like this: > + > + SET character_set_collations='utf8mb3=utf8mb3_bin'; > + INSERT INTO t1 VALUES (_utf8mb3'test'); > + COMMIT; > + > + The statment (INSERT in this example) is already in binlog at this point, and the > + and the "SET character_set_collations" is written inside a > + Q_CHARACTER_SET_COLLATIONS chunk in its log entry header. > + The flag CHARACTER_SET_COLLATIONS_USED is not needed any more. > + > + Let's suppress the flag to avoid a Q_CHARACTER_SET_COLLATIONS chunk > + inside the COMMIT log entry header - it would be useless and would > + only waste space in the binary log. > + */ > + thd->used&= ~THD::CHARACTER_SET_COLLATIONS_USED; Wouldn't that logic apply to other THD::xxx_USED flags? May be you should reset thd->used=0 here? > + > if (unlikely(thd->is_error()) || > (thd->variables.option_bits & OPTION_MASTER_SQL_ERROR)) > { > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -12457,8 +12459,10 @@ bool HA_CREATE_INFO:: I'd prefer you didn't put a line break after :: see how useful it makes the hunk header? > { > // Make sure we don't do double resolution in direct SQL execution > DBUG_ASSERT(!default_table_charset || thd->stmt_arena->is_stmt_execute()); > + Character_set_collations_used used(thd); > if (!(default_table_charset= > - default_cscl.resolved_to_context(ctx))) > + default_cscl.resolved_to_context(&used, > + thd->variables.character_set_collations, ctx))) > return true; > } > > diff --git a/sql/sql_class.h b/sql/sql_class.h > --- a/sql/sql_class.h > +++ b/sql/sql_class.h > @@ -5620,6 +5624,29 @@ class THD: public THD_count, /* this must be first */ > }; > > > +class Character_set_collations_used: public Charset_collation_map_st::Used > +{ > + THD *m_thd; > +public: > + Character_set_collations_used(THD *thd) > + :m_thd(thd) > + { } > + ~Character_set_collations_used() > + { > + /* > + Mark THD that the collation map was used, > + no matter if a compiled or a mapped collation was why no matter? as far as I understand, you didn't plan to change compiled-in defaults > + found during charset->collation resolution. > + Even if the map was empty, we still need to print > + SET @@session.character_set_collations=''; > + in mariadb-binlog output. > + */ > + if (m_used) > + m_thd->used|= THD::CHARACTER_SET_COLLATIONS_USED; > + } > +}; > + > + > /* > Start a new independent transaction for the THD. > The old one is stored in this object and restored when calling > diff --git a/sql/sql_prepare.cc b/sql/sql_prepare.cc > --- a/sql/sql_prepare.cc > +++ b/sql/sql_prepare.cc > @@ -3538,6 +3544,13 @@ static void mysql_stmt_execute_common(THD *thd, > DBUG_VOID_RETURN; > } > > + if (stmt->prepare_time_charset_collation_map_version() != > + thd->variables.character_set_collations.version()) > + { > + if (stmt->reprepare()) > + DBUG_VOID_RETURN; > + } why do you reprepare here and now where reprepare() is normally happens, in the Prepared_statement::execute_loop()? > + > /* > In case of direct execution application decides how many parameters > to send. > @@ -5220,6 +5252,13 @@ bool Prepared_statement::execute(String *expanded_query, bool open_cursor) > MYSQL_QUERY_EXEC_START(thd->query(), thd->thread_id, thd->get_db(), > &thd->security_ctx->priv_user[0], > (char *) thd->security_ctx->host_or_ip, 1); > + /* > + If PREPARE used @@character_set_collations, > + then we need to make sure binary log writes > + the map in the event header. > + */ why is it different from other THD::xx_USED flags? > + thd->used|= m_prepare_time_thd_used_flags & > + THD::CHARACTER_SET_COLLATIONS_USED; > error= mysql_execute_command(thd, true); > MYSQL_QUERY_EXEC_DONE(error); > thd->update_server_status(); > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -436,6 +436,115 @@ static bool update_auto_increment_increment (sys_var *self, THD *thd, enum_var_t > > #endif /* WITH_WSREP */ > > + > +class Sys_var_charset_collation_map: public sys_var in sys_var.inl, please > +{ > +public: > + Sys_var_charset_collation_map(const char *name_arg, const char *comment, > + int flag_args, ptrdiff_t off, size_t size, > + CMD_LINE getopt, > + enum binlog_status_enum binlog_status_arg) > + :sys_var(&all_sys_vars, name_arg, comment, > + flag_args, off, getopt.id, getopt.arg_type, > + SHOW_CHAR, > + DEFAULT(0), nullptr, binlog_status_arg, > + nullptr, nullptr, nullptr) > + { > + option.var_type|= GET_STR; > + } > + > +private: > + > + static bool charset_collation_map_from_item(Charset_collation_map_st *map, > + Item *item, > + myf utf8_flag) > + { > + String *value, buffer; > + if (!(value= item->val_str_ascii(&buffer))) > + return true; > + return map->from_text(value->to_lex_cstring(), utf8_flag); > + } > + > + static const uchar *make_value_ptr(THD *thd, > + const Charset_collation_map_st &map) > + { > + size_t nbytes= map.text_format_nbytes_needed(); > + char *buf= (char *) thd->alloc(nbytes); > + size_t length= map.print(buf, nbytes); > + return (uchar *) thd->strmake(buf, length); Eh, sorry. I don't understand. You allocate memory twice? for two copies of the string? > + } > + > + > +private: > + > + bool do_check(THD *thd, set_var *var) override > + { > + Charset_collation_map_st map; > + return charset_collation_map_from_item(&map, var->value, > + thd->get_utf8_flag()); > + } > + > + void session_save_default(THD *thd, set_var *var) override > + { > + thd->variables.character_set_collations.set( > + global_system_variables.character_set_collations, 1); > + } > + > + void global_save_default(THD *thd, set_var *var) override > + { > + global_system_variables.character_set_collations.init(); > + } > + > + bool session_update(THD *thd, set_var *var) override > + { > + Charset_collation_map_st map; > + if (!var->value) > + { > + session_save_default(thd, var); > + return false; > + } > + if (charset_collation_map_from_item(&map, var->value, thd->get_utf8_flag())) The idea is that var->value is only evaluated in check and is not reevaluated in update. check is supposed to store the value in set_var *var for update to use. > + return true; > + thd->variables.character_set_collations.set(map, 1); > + return false; > + } > + > + bool global_update(THD *thd, set_var *var) override > + { > + Charset_collation_map_st map; > + if (!var->value) > + { > + global_save_default(thd, var); > + return false; > + } > + if (charset_collation_map_from_item(&map, var->value, thd->get_utf8_flag())) > + return true; > + global_system_variables.character_set_collations= map; > + return false; > + } > + > + const uchar * > + session_value_ptr(THD *thd, const LEX_CSTRING *base) const override > + { > + return make_value_ptr(thd, thd->variables.character_set_collations); > + } > + > + const uchar * > + global_value_ptr(THD *thd, const LEX_CSTRING *base) const override > + { > + return make_value_ptr(thd, global_system_variables. > + character_set_collations); > + } > +}; > + > + > +static Sys_var_charset_collation_map Sys_character_set_collations( > + "character_set_collations", > + "Default collations for character sets", > + SESSION_VAR(character_set_collations), > + NO_CMD_LINE, NOT_IN_BINLOG); > + > + > static Sys_var_double Sys_analyze_sample_percentage( > "analyze_sample_percentage", > "Percentage of rows from the table ANALYZE TABLE will sample " > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy > --- a/sql/sql_yacc.yy > +++ b/sql/sql_yacc.yy > @@ -9686,6 +9713,9 @@ column_default_non_parenthesized_expr: > } > | CONVERT_SYM '(' expr USING charset_name ')' > { > + Character_set_collations_used used(thd); > + $5= thd->variables.character_set_collations. > + get_collation_for_charset(&used, $5); this is rather awkward pattern that you use everywhere, with Character_set_collations_used used(thd); why not to pass thd as an argument instead? - Character_set_collations_used used(thd); $5= thd->variables.character_set_collations. - get_collation_for_charset(&used, $5); + get_collation_for_charset(thd, $5); > $$= new (thd->mem_root) Item_func_conv_charset(thd, $3, $5); > if (unlikely($$ == NULL)) > MYSQL_YYABORT; > diff --git a/sql/simple_tokenizer.h b/sql/simple_tokenizer.h > --- /dev/null > +++ b/sql/simple_tokenizer.h > @@ -0,0 +1,83 @@ > +/* Copyright (c) 2022, MariaDB Corporation. 2023 > + > + This program is free software; you can redistribute it and/or modify > + it under the terms of the GNU General Public License as published by > + the Free Software Foundation; version 2 of the License. > + > + This program is distributed in the hope that it will be useful, > + but WITHOUT ANY WARRANTY; without even the implied warranty of > + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + GNU General Public License for more details. > + > + You should have received a copy of the GNU General Public License > + along with this program; if not, write to the Free Software > + Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1335 USA */ > + > +#ifndef SIMPLE_TOKENIZER_INCLUDED > +#define SIMPLE_TOKENIZER_INCLUDED > + > + > +class Simple_tokenizer > +{ > + const char *m_ptr; > + const char *m_end; > +public: > + Simple_tokenizer(const char *str, size_t length) > + :m_ptr(str), m_end(str + length) > + { } > + const char *ptr() const > + { > + return m_ptr; > + } > + bool eof() const > + { > + return m_ptr >= m_end; > + } > + void get_spaces() > + { > + for ( ; !eof(); m_ptr++) > + { > + if (m_ptr[0] != ' ') > + break; > + } > + } > + bool is_ident_start(char ch) const > + { > + return (ch >= 'a' && ch <= 'z') || > + (ch >= 'A' && ch <= 'Z') || > + ch == '_'; > + } > + bool is_ident_body(char ch) const > + { > + return is_ident_start(ch) || > + (ch >= '0' && ch <= '9'); > + } > + bool is_ident_start() const > + { > + return !eof() && is_ident_start(*m_ptr); > + } > + bool is_ident_body() const > + { > + return !eof() && is_ident_body(*m_ptr); > + } > + LEX_CSTRING get_ident() > + { > + if (!is_ident_start()) > + return {m_ptr,0}; > + const char *start= m_ptr++; > + for ( ; is_ident_body(); m_ptr++) > + { } > + LEX_CSTRING res= {start, (size_t) (m_ptr - start)}; > + return res; > + } > + bool get_char(char ch) > + { > + if (eof() || *m_ptr != ch) > + return true; > + m_ptr++; > + return false; > + } > +}; that's too big and complex, but also not complex enough. note that we have quite a few of those already, in patricular in typelib.c, in strfunc.cc, item_func.cc (find_in_set) etc. and this one could've been 2-3 times smaller for @@character_set_collations only, and it still cannot replace any other parsers from above. I'd keep just two methods: get_token() - which skips whitespaces and returns the next substring of non-white-space and not ',' or '=' characters. multibyte-safe. And get_char, which returns the next non-white-space character. this parser could be used in all use cases from above. > + > + > +#endif // SIMPLE_TOKENIZER_INCLUDED > diff --git a/sql/charset_collations.h b/sql/charset_collations.h > --- /dev/null > +++ b/sql/charset_collations.h > @@ -0,0 +1,265 @@ > +/* Copyright (c) 2023, MariaDB Corporation. > + > + This program is free software; you can redistribute it and/or modify > + it under the terms of the GNU General Public License as published by > + the Free Software Foundation; version 2 of the License. > + > + This program is distributed in the hope that it will be useful, > + but WITHOUT ANY WARRANTY; without even the implied warranty of > + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + GNU General Public License for more details. > + > + You should have received a copy of the GNU General Public License > + along with this program; if not, write to the Free Software > + Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1335 USA */ > + > +#ifndef LEX_CHARSET_COLLATIONS_INCLUDED > +#define LEX_CHARSET_COLLATIONS_INCLUDED > + > +struct Charset_collation_map_st > +{ > +public: > + > + class Used > + { > + public: > + enum map_used_t > + { > + USED_NONE= 0, > + USED_COMPILED_COLLATION= 1 << 0, > + USED_MAPPED_COLLATION= 1 << 1 > + }; > + protected: > + map_used_t m_used; > + public: > + Used() > + :m_used(USED_NONE) > + { } > + void add(map_used_t flag) > + { > + m_used= (map_used_t) ((uint) m_used | (uint) flag); > + } > + }; > + > + struct Elem_st > + { > + protected: > + CHARSET_INFO *m_charset; > + CHARSET_INFO *m_collation; What is the difference between charset and collation? I was under impression that any CHARSET_INFO describes a collation, there is no special structure just for a charset. Am I wrong? > + static size_t print_lex_string(char *dst, const LEX_CSTRING &str) > + { > + memcpy(dst, str.str, str.length); > + return str.length; > + } > + public: > + /* > + Size in text format: 'utf8mb4=utf8mb4_unicode_ai_ci' > + */ > + static constexpr size_t text_size_max() > + { > + return MY_CS_CHARACTER_SET_NAME_SIZE + 1 + > + MY_CS_COLLATION_NAME_SIZE; > + } > + CHARSET_INFO *charset() const > + { > + return m_charset; > + } > + CHARSET_INFO *collation() const > + { > + return m_collation; > + } > + void set_collation(CHARSET_INFO *cl) > + { > + m_collation= cl; > + } > + size_t print(char *dst) const > + { > + const char *dst0= dst; > + dst+= print_lex_string(dst, m_charset->cs_name); > + *dst++= '='; > + dst+= print_lex_string(dst, m_collation->coll_name); > + return (size_t) (dst - dst0); > + } > + int cmp_by_charset_id(const Elem_st &rhs) const > + { > + return m_charset->number < rhs.m_charset->number ? -1 : > + m_charset->number > rhs.m_charset->number ? +1 : 0; > + } > + }; > + class Elem: public Elem_st > + { > + public: > + Elem(CHARSET_INFO *charset, CHARSET_INFO *collation) > + { > + m_charset= charset; > + m_collation= collation; > + } > + }; > +protected: > + Elem_st m_element[8]; // Should be enough for now > + uint m_count; > + uint m_version; > + > + static int cmp_by_charset_id(const void *a, const void *b) > + { > + return static_cast<const Elem_st*>(a)-> > + cmp_by_charset_id(*static_cast<const Elem_st*>(b)); > + } > + > + void sort() > + { > + qsort(m_element, m_count, sizeof(Elem_st), cmp_by_charset_id); > + } > + > + const Elem_st *find_elem_by_charset_id(uint id) const > + { > + if (!m_count) > + return NULL; > + int first= 0, last= ((int) m_count) - 1; > + for ( ; first <= last; ) > + { > + const int middle= (first + last) / 2; > + DBUG_ASSERT(middle >= 0); > + DBUG_ASSERT(middle < (int) m_count); > + const uint middle_id= m_element[middle].charset()->number; > + if (middle_id == id) > + return &m_element[middle]; > + if (middle_id < id) > + first= middle + 1; > + else > + last= middle - 1; > + } > + return NULL; > + } > + > + bool insert(const Elem_st &elem) > + { > + DBUG_ASSERT(elem.charset()->state & MY_CS_PRIMARY); > + if (m_count >= array_elements(m_element)) > + return true; > + m_element[m_count]= elem; > + m_count++; > + sort(); > + return false; > + } > + > + bool insert_or_replace(const Elem_st &elem) > + { > + DBUG_ASSERT(elem.charset()->state & MY_CS_PRIMARY); > + const Elem_st *found= find_elem_by_charset_id(elem.charset()->number); > + if (found) > + { > + const_cast<Elem_st*>(found)->set_collation(elem.collation()); > + return false; > + } > + return insert(elem); > + } > + > +public: > + void init() > + { > + m_count= 0; > + m_version= 0; > + } > + uint count() const > + { > + return m_count; > + } > + uint version() const > + { > + return m_version; > + } > + void set(const Charset_collation_map_st &rhs, uint version_increment) > + { > + uint version= m_version; > + *this= rhs; > + m_version= version + version_increment; > + } > + const Elem_st & operator[](uint pos) const > + { > + DBUG_ASSERT(pos < m_count); > + return m_element[pos]; > + } > + bool insert_or_replace(const class Lex_exact_charset &cs, > + const class Lex_extended_collation &cl, > + bool error_on_conflicting_duplicate); > + CHARSET_INFO *get_collation_for_charset(Used *used, > + CHARSET_INFO *cs) const > + { > + DBUG_ASSERT(cs->state & MY_CS_PRIMARY); > + const Elem_st *elem= find_elem_by_charset_id(cs->number); > + if (elem) > + { > + used->add(Used::USED_MAPPED_COLLATION); > + return elem->collation(); > + } > + used->add(Used::USED_COMPILED_COLLATION); > + return cs; > + } > + size_t text_format_nbytes_needed() const > + { > + return (Elem_st::text_size_max() + 1/* for ',' */) * m_count; > + } > + size_t print(char *dst, size_t nbytes_available) const > + { > + const char *dst0= dst; > + const char *end= dst + nbytes_available; > + for (uint i= 0; i < m_count; i++) > + { > + if (Elem_st::text_size_max() + 1/* for ',' */ > (size_t) (end - dst)) > + break; > + if (i > 0) > + *dst++= ','; > + dst+= m_element[i].print(dst); > + } > + return dst - dst0; > + } > + static constexpr size_t binary_size_max() > + { > + return 1/*count*/ + 4 * array_elements(m_element); > + } > + size_t to_binary(char *dst) const > + { > + const char *dst0= dst; > + *dst++= (char) (uchar) m_count; > + for (uint i= 0; i < m_count; i++) > + { > + int2store(dst, (uint16) m_element[i].charset()->number); > + dst+= 2; > + int2store(dst, (uint16) m_element[i].collation()->number); > + dst+= 2; > + } > + return (size_t) (dst - dst0); > + } > + size_t from_binary(const char *src, size_t srclen) > + { > + const char *src0= src; > + init(); > + if (!srclen) > + return 0; // Empty > + uint count= (uchar) *src++; > + if (srclen < 1 + 4 * count) > + return 0; > + for (uint i= 0; i < count; i++, src+= 4) > + { > + CHARSET_INFO *cs, *cl; > + if (!(cs= get_charset(uint2korr(src), MYF(0))) || > + !(cl= get_charset(uint2korr(src + 2), MYF(0)))) > + { > + /* > + Unpacking from binary format happens on the slave side. > + If for some reasons the slave does not know about a > + character set or a collation, just skip the pair here. > + This pair might not even be needed. > + */ > + continue; > + } > + insert_or_replace(Elem(cs, cl)); > + } > + return src - src0; > + } > + bool from_text(const LEX_CSTRING &str, myf utf8_flag); > +}; > + > + > +#endif // LEX_CHARSET_COLLATIONS_INCLUDED Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 3

Re: d8e04ef367b: MDEV-25237 Assertion `global_system_variables.session_track_system_variables' failed in Session_sysvars_tracker::init | SIGSEGV's in __strlen_avx2 | UBSAN: runtime error: null pointer passed as argument 1, which is declared to never be null in my_strdup
by Sergei Golubchik 07 Jul '23

07 Jul '23

Hi, Oleksandr, On Jul 07, Oleksandr Byelkin wrote: > revision-id: d8e04ef367b (mariadb-10.4.30-9-gd8e04ef367b) > parent(s): 423c28f0aa4 > author: Oleksandr Byelkin > committer: Oleksandr Byelkin > timestamp: 2023-07-03 17:50:49 +0200 > message: > > MDEV-25237 Assertion `global_system_variables.session_track_system_variables' failed in Session_sysvars_tracker::init | SIGSEGV's in __strlen_avx2 | UBSAN: runtime error: null pointer passed as argument 1, which is declared to never be null in my_strdup first, if you see a bug with a very long summary, particularly if it's basically makes no sense to a user - in that case, please, don't hesitate to make it shorter and more meaningful. For example "crash after setting session_track_system_variables to an invalid value" > --- a/sql/sys_vars.inl > +++ b/sql/sys_vars.inl > @@ -620,7 +620,12 @@ class Sys_var_sesvartrack: public Sys_var_charptr_base > { > if (sysvartrack_global_update(thd, new_val, > var->save_result.string_value.length)) > - new_val= 0; > + { > + if (new_val) > + my_free(new_val); > + // keep the old value in case of error > + new_val= (char*)my_strdup(global_var(const char*), MYF(MY_WME)); > + } No, this is wrong. Well, not wrong, but an incorrect bug fix. update should not fail, this is the contract. check() is supposed to verify that the value is valid, so basically update must succeed. technically it can fail because of OOM, I suppose, so restoring the value here might be needed still. But in the case of OOM, strdup() might be not such a great idea. Perhaps you can make sure that OOM is impossible and update truly cannot fail? Anyway, the actual fix for MDEV-25237 is to make check fail on an invalid value. In fact, it was a simple typo: @@ -221,7 +221,7 @@ bool sysvartrack_validate_value(THD *thd, const char *str, > /* Remove leading/trailing whitespace. */ trim_whitespace(system_charset_info, &var); - if (!strcmp(var.str, "*") && !find_sys_var(thd, var.str, var.length)) + if (strcmp(var.str, "*") && !find_sys_var(thd, var.str, var.length)) return true; > } > global_update_finish(new_val); > return (new_val == 0 && var->save_result.string_value.str != 0); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: d23555b1920: MDEV-30984 Online ALTER table is denied with non-informative error messages
by Sergei Golubchik 06 Jul '23

06 Jul '23

Hi, Nikita, On Jun 24, Nikita Malyavin wrote: > revision-id: d23555b1920 (mariadb-11.0.1-138-gd23555b1920) > parent(s): df90bf34644 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-06-20 16:10:48 +0300 > message: > > MDEV-30984 Online ALTER table is denied with non-informative error messages > > diff --git a/mysql-test/main/alter_table_online.result b/mysql-test/main/alter_table_online.result > --- a/mysql-test/main/alter_table_online.result > +++ b/mysql-test/main/alter_table_online.result > @@ -179,25 +179,27 @@ t CREATE TABLE `t` ( > UNIQUE KEY `b` (`b`) > ) ENGINE=InnoDB DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci > alter table t drop b, change c c serial, algorithm=copy, lock=none; > -ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > +ERROR 0A000: LOCK=NONE is not supported. Reason: CHANGE COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > # Check existed unique keys. > create or replace table t(a int, b int not null, c int not null, d int); > # No unique in the old table; > alter table t add unique(b, c), modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > -ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > +ERROR 0A000: LOCK=NONE is not supported. Reason: CHANGE COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > alter table t add unique(a, b); > # Unique in the old table has nulls; > alter table t modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > -ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > +ERROR 0A000: LOCK=NONE is not supported. Reason: CHANGE COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > alter table t add unique(b, c); > # Change unique'scolumn; Typo > -alter table t change b x int, modify d int auto_increment, add key(d), > +alter table t change b x bigint, modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > -ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > -# Finally good. > -alter table t modify d int auto_increment, add key(d), > +ERROR 0A000: LOCK=NONE is not supported. Reason: CHANGE COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > +# Finally good. Simple renames with a type unchenged will not affect typo s/unchenged/unchanged/ > +# the result. Also NOT NULL -> NULL transform is fine. > +alter table t modify d int auto_increment, add key(d), > +change b x int null, > algorithm=copy, lock=none; > drop table t; > # MDEV-31172 Server crash or ASAN errors in online_alter_check_autoinc > diff --git a/sql/sql_class.h b/sql/sql_class.h > --- a/sql/sql_class.h > +++ b/sql/sql_class.h > @@ -1228,6 +1228,15 @@ class Query_arena > { return strdup_root(mem_root,str); } > inline char *strmake(const char *str, size_t size) const > { return strmake_root(mem_root,str,size); } > + inline LEX_CSTRING strcat(const LEX_CSTRING &a, const LEX_CSTRING &b) const > + { > + char *buf= (char*)alloc(a.length + b.length + 1); > + if (unlikely(!buf)) > + return null_clex_str; > + strncpy(buf, a.str, a.length); memcpy both ↑ and ↓ (and then you might need to append \0 explicitly) > + strncpy(buf + a.length, b.str, b.length + 1); > + return {buf, a.length + b.length}; > + } > inline void *memdup(const void *str, size_t size) const > { return memdup_root(mem_root,str,size); } > inline void *memdup_w_gap(const void *str, size_t size, size_t gap) const > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -9869,26 +9869,21 @@ bool online_alter_check_autoinc(const THD *thd, const Alter_info *alter_info, > */ > for (uint k= 0; k < table->s->keys; k++) > { > - const KEY &key= table->s->key_info[k]; > + const KEY &key= table->key_info[k]; This is generally incorrect. User specified keys are in table->s->key_info[] Keys in table->key_info[] can be modified to match what indexes are actually created, in particular, for long uniques table->key_info[] differs from table->s->key_info[] > if ((key.flags & HA_NOSAME) == 0 || key.flags & HA_NULL_PART_KEY) > continue; > bool key_parts_good= true; > for (uint kp= 0; kp < key.user_defined_key_parts && key_parts_good; kp++) > { > - const char *field_name= key.key_part[kp].field->field_name.str; > - for (const auto &c: alter_info->drop_list) > - if (c.type == Alter_drop::COLUMN > - && my_strcasecmp(system_charset_info, c.name, field_name) == 0) > - { > - key_parts_good= false; > - break; > - } > + const Field *f= key.key_part[kp].field; > + // tmp_set contains dropped fields after mysql_prepare_alter_table > + key_parts_good= !bitmap_is_set(&table->tmp_set, f->field_index); > + > if (key_parts_good) > for (const auto &c: alter_info->create_list) > - if (c.change.str && my_strcasecmp(system_charset_info, c.change.str, > - field_name) == 0) > + if (c.field == f) > { > - key_parts_good= false; > + key_parts_good= f->is_equal(c); > break; > } > } > @@ -9896,25 +9891,63 @@ bool online_alter_check_autoinc(const THD *thd, const Alter_info *alter_info, > return true; > } > > - const char *old_autoinc= table->found_next_number_field > - ? table->found_next_number_field->field_name.str > - : ""; > - bool online= true; > for (const auto &c: alter_info->create_list) > { > - if (c.change.str && c.flags & AUTO_INCREMENT_FLAG) > + if (c.flags & AUTO_INCREMENT_FLAG) > { > - if (my_strcasecmp(system_charset_info, c.change.str, old_autoinc) != 0) > + if (c.field && !(c.field->flags & AUTO_INCREMENT_FLAG)) > + return false; > + break; > + } > + } > + return true; > +} > + > +static > +const char *online_alter_check_supported(const THD *thd, > + const Alter_info *alter_info, > + const TABLE *table, bool *online) > { > - if (c.create_if_not_exists // check IF EXISTS option > - && table->find_field_by_name(&c.change) == NULL) > - continue; > - online= false; > + DBUG_ASSERT(!table->s->tmp_table); that's confusing. *online can be false here already, then return "DROP SYSTEM VERSIONING" will be wrong and you use some other condition to decide whether "DROP SYSTEM VERSIONING" is correct or not. it'd be much clearer not to call online_alter_check_supported in this case. and add here DBUG_ASSERT(*online); > + > + *online= *online && (alter_info->flags & ALTER_DROP_SYSTEM_VERSIONING) == 0; and then you won't need to use && above, just as you don't use it below > + if (!*online) > + return "DROP SYSTEM VERSIONING"; > + > + *online= !thd->lex->ignore; > + if (!*online) > + return "ALTER IGNORE TABLE"; > + > + *online= !table->versioned(VERS_TRX_ID); > + if (!*online) > + return "BIGINT GENERATED ALWAYS AS ROW_START"; > + > + List<FOREIGN_KEY_INFO> fk_list; > + table->file->get_foreign_key_list(thd, &fk_list); > + for (auto &fk: fk_list) > + { > + if (fk_modifies_child(fk.delete_method) || > + fk_modifies_child(fk.update_method)) > + { > + *online= false; > + // Don't fall to a common unsupported case to avoid heavy string ops. > + if (alter_info->requested_lock == Alter_info::ALTER_TABLE_LOCK_NONE) > + { > + return fk_modifies_child(fk.delete_method) > + ? thd->strcat({STRING_WITH_LEN("ON DELETE ")}, > + *fk_option_name(fk.delete_method)).str > + : thd->strcat({STRING_WITH_LEN("ON UPDATE ")}, > + *fk_option_name(fk.update_method)).str; > } > - break; > + return NULL; > } > } > - return online; > + > + *online= online_alter_check_autoinc(thd, alter_info, table); > + if (!*online) > + return "CHANGE COLUMN ... AUTO_INCREMENT"; > + > + return NULL; > } > > > @@ -10979,14 +10994,14 @@ do_continue:; > > if (!table->s->tmp_table) > { > - // COPY algorithm doesn't work with concurrent writes. > + auto *reason= online_alter_check_supported(thd, alter_info, table, &online); > + // COPY algorithm works with concurrent writes only when online is true. as I wrote above, better put it inside if (online) { > if (!online && > alter_info->requested_lock == Alter_info::ALTER_TABLE_LOCK_NONE) > { > + DBUG_ASSERT(reason); > my_error(ER_ALTER_OPERATION_NOT_SUPPORTED_REASON, MYF(0), > - "LOCK=NONE", > - ER_THD(thd, ER_ALTER_OPERATION_NOT_SUPPORTED_REASON_COPY), > - "LOCK=SHARED"); > + "LOCK=NONE", reason, "LOCK=SHARED"); > goto err_new_table_cleanup; > } > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

3 7

[MariaDB developers]Re: a2e71bc8e89: MDEV-22979 MDEV-31400 Fixing spider init bugs
by Yuchen Pei 03 Jul '23

03 Jul '23

Hi Sergei, Thanks for the comments. I have addressed them and the url of the new patch can be found in the new request for review comment in the ticket. On Mon 2023-06-12 18:07:03 +0200, Sergei Golubchik wrote: > Hi, Yuchen, > > See comments inline. Here I'm only reviewing MDEV-31400, > please, tell me if you'd like me to review spider part too > Given that Alexey (cc'd) has already been working on the review of the spider init bug fix, I think it makes sense to ask Alexey to continue on the spider part, which I just did on MDEV-22979. Given that the spider part requires the patch for MDEV-31400 to work, I have the spider part on top of the init dependency part. > On Jun 12, Yuchen Pei wrote: >> revision-id: a2e71bc8e89 (mariadb-10.9.5-19-ga2e71bc8e89) >> parent(s): cb0e0c915f6 >> author: Yuchen Pei >> committer: Yuchen Pei >> timestamp: 2023-06-08 16:13:56 +1000 >> message: >> >> MDEV-22979 MDEV-31400 Fixing spider init bugs > > please, make a separate commit for MDEV-31400 Done. > >> diff --git a/sql/handler.cc b/sql/handler.cc >> index a12e9ea18f5..60080f1da6a 100644 >> --- a/sql/handler.cc >> +++ b/sql/handler.cc >> @@ -646,10 +648,14 @@ int ha_initialize_handlerton(st_plugin_int *plugin) >> hton->slot= HA_SLOT_UNDEF; >> /* Historical Requirement */ >> plugin->data= hton; // shortcut for the future >> - if (plugin->plugin->init && plugin->plugin->init(hton)) >> + if (plugin->plugin->init && (ret= plugin->plugin->init(hton))) >> { >> - sql_print_error("Plugin '%s' init function returned error.", >> - plugin->name.str); >> + if (unlikely(ret == -1)) >> + sql_print_warning("Plugin '%s' init function returned error but >> may be retried.", >> + plugin->name.str); >> + else >> + sql_print_error("Plugin '%s' init function returned error.", >> + plugin->name.str); > > let's treat all plugins identically. Meaning, in particular, no warnings > or errors in ha_initialize_handlerton(). They belong to sql_plugin.cc > Done. >> goto err; >> } >> >> diff --git a/sql/mysqld.cc b/sql/mysqld.cc >> index 32392ab882e..caf85736770 100644 >> --- a/sql/mysqld.cc >> +++ b/sql/mysqld.cc >> @@ -365,6 +365,7 @@ bool opt_disable_networking=0, opt_skip_show_db=0; >> bool opt_skip_name_resolve=0; >> my_bool opt_character_set_client_handshake= 1; >> bool opt_endinfo, using_udf_functions; >> +bool udf_initialized= 0; > > I don't think you need that. You can use exising mysqld_server_started. > That won't work, because mysqld_server_started will only be turned on very late, towards the end of mysqld_main(). If spider is initialised between udf_init() and mysqld_server_started is 1 (for example in an init-file), then the udf will fail. See https://github.com/MariaDB/server/commit/91176a7be7c where the test udf_mysql_func_early fails with "query 'SELECT SPIDER_DIRECT_SQL('select * from tbl_a', 'results', 'srv "s_2_1", database "auto_test_remote"')' failed: ER_SP_DOES_NOT_EXIST (1305): FUNCTION auto_test_local.SPIDER_DIRECT_SQL does not exist" because of this. Now, I assume MDEV-31401, the task to include the udf insertion trick in early calls CREATE FUNCTION, will fix this issue without introduce any new global booleans because it should has access to the internal variable `initialized`, though I do not know how complex or doable this task is. So I guess there's a trade-off here between fixing the spider init bugs now and investigating MDEV-31401. Would you like me to explore MDEV-31401 first? >> my_bool locked_in_memory; >> bool opt_using_transactions; >> bool volatile abort_loop; >> diff --git a/sql/sql_plugin.cc b/sql/sql_plugin.cc >> index 5a077a934ac..e564ab5a38d 100644 >> --- a/sql/sql_plugin.cc >> +++ b/sql/sql_plugin.cc >> @@ -1462,10 +1462,16 @@ static int plugin_initialize(MEM_ROOT *tmp_root, struct st_plugin_int *plugin, >> >> if (plugin_type_initialize[plugin->plugin->type]) >> { >> - if ((*plugin_type_initialize[plugin->plugin->type])(plugin)) >> + ret= (*plugin_type_initialize[plugin->plugin->type])(plugin); >> + if (ret) >> { >> - sql_print_error("Plugin '%s' registration as a %s failed.", >> - plugin->name.str, plugin_type_names[plugin->plugin->type].str); >> + /* Plugin init failed but requested a retry if possible */ >> + if (unlikely(ret == -1)) >> + sql_print_warning("Plugin '%s' registration as a %s failed but may be retried.", >> + plugin->name.str, plugin_type_names[plugin->plugin->type].str); >> + else >> + sql_print_error("Plugin '%s' registration as a %s failed.", >> + plugin->name.str, plugin_type_names[plugin->plugin->type].str); > > 1. "may be" is confusing, we should be less vague in the error > messages Done. I have expanded it a bit more to the following but we still really use "will be" here without either losing accuracy or being overly verbose. Even if it fails and returns the retry code, the retry will not happen if no new plugins are successfully initialised this round. --8<---------------cut here---------------start------------->8--- sql_print_warning("Plugin '%s' registration as a %s failed and may " "be retried provided it is being loaded with " "plugin-load[-add]", plugin->name.str, plugin_type_names[plugin->plugin->type].str); --8<---------------cut here---------------end--------------->8--- > > 2. about (ret == -1). I looked at what plugin init functions return now: > 33 plugins don't have an init function > 76 plugins always return 0 > 13 plugins return 0 or 1 > 3 plugins return 0 or -1 > 3 plugins return 0 or HA_ERR_INITIALIZATION, HA_ERR_OUT_OF_MEM > and innodb inconsistently return 0, HA_ERR_xxx as above, or 1 > > so, I suggest to introduce, like, HA_ERR_DEPENDENCIES (or whatever) > and only retry plugins that return that specific error code. > Done. The HA_ERR_ namespace is tightly packed and bounded by HA_ERR_FIRST (120) and HA_ERR_LAST (198), and I crammed in a #define HA_ERR_RETRY_INIT 129. >> goto err; >> } >> } >> @@ -1462,8 +1462,14 @@ static int plugin_initialize(MEM_ROOT *tmp_root, struct st_plugin_int *plugin, >> >> if (plugin_type_initialize[plugin->plugin->type]) >> { >> - if ((*plugin_type_initialize[plugin->plugin->type])(plugin)) >> + ret= (*plugin_type_initialize[plugin->plugin->type])(plugin); > > there's a second branch below, with plugin->plugin->init(), > please, handle it too. May be, do as with deinit? > > plugin_type_init deinit= plugin_type_deinitialize[plugin->plugin->type]; > if (!deinit) > deinit= (plugin_type_init)(plugin->plugin->deinit); > > if (deinit && deinit(plugin)) > > Less code duplication here. > Done. >> + if (ret) >> { >> + /* Plugin init failed but requested a retry if possible */ >> + if (unlikely(ret == -1)) >> + sql_print_warning("Plugin '%s' registration as a %s failed but may be retried.", >> + plugin->name.str, plugin_type_names[plugin->plugin->type].str); >> + else >> sql_print_error("Plugin '%s' registration as a %s failed.", >> plugin->name.str, plugin_type_names[plugin->plugin->type].str); >> goto err; >> @@ -1737,11 +1743,34 @@ int plugin_init(int *argc, char **argv, int flags) >> */ >> >> mysql_mutex_lock(&LOCK_plugin); >> + /* List of plugins to reap */ >> reap= (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); >> *(reap++)= NULL; >> + /* List of plugins to retry */ >> + retry= (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); >> + *(retry++)= NULL; >> >> for(;;) >> { >> + /* Number of plugins that is successfully initialised in a round */ >> + int num_initialized; >> + do >> + { >> + num_initialized= 0; >> + /* If any plugins failed and requested a retry, clean up before >> + retry */ >> + while ((plugin_ptr= *(--retry))) >> + { >> + mysql_mutex_unlock(&LOCK_plugin); >> + plugin_deinitialize(plugin_ptr, true); >> + mysql_mutex_lock(&LOCK_plugin); >> + /** Needed to satisfy assertions in `test_plugin_options()` */ > > Hmm, I don't think it'll work. I mean, test_plugin_options() - as far as > I remember - will remove processed options from the command line array, > meaning you cannot call it twice for the same plugin. > > May be it'd be better to repeat only plugin init function call and not > the complete initialize/deinitialize cycle? Done. I separated out a plugin_do_initialize() function that skips the test_options() part and largely mirrors the plugin_deinitialize() function, and calls this function instead of the full plugin_initialize() during a retry. I also made sure the code path outside of plugin_init() does not change, i.e. if plugin_initialize() is called from an INSTALL statement. The contract of init and deinit is not clear, but I suppose it is a reasonable expectation that if a plugin init wants to be retried, it will clean up before returning init instead of relying the server to call the deinit before retrying. > >> + my_afree(plugin_ptr->ptr_backup); >> + plugin_ptr->ptr_backup= NULL; >> + plugin_ptr->nbackups= 0; >> + plugin_ptr->state= PLUGIN_IS_UNINITIALIZED; >> + } >> + retry++; >> for (i=0; i < MYSQL_MAX_PLUGIN_TYPE_NUM; i++) >> { >> HASH *hash= plugin_hash + plugin_type_initialization_order[i]; > > Regards, > Sergei > VP of MariaDB Server Engineering > and security(a)mariadb.org Best, Yuchen

2 3

Re: 8b5de389ab1: MDEV-31400 Simple plugin dependency resolution
by Sergei Golubchik 03 Jul '23

03 Jul '23

Hi, Yuchen, On Jul 01, Yuchen Pei wrote: > revision-id: 8b5de389ab1 (mariadb-10.9.5-17-g8b5de389ab1) > parent(s): d8997f875e2 > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-06-13 20:08:28 +1000 > message: > > MDEV-31400 Simple plugin dependency resolution > > We introduce simple plugin dependency. A plugin init function may > return HA_ERR_RETRY_INIT. If this happens during server startup when > the server is trying to initialise all plugins, the failed plugins > will be retried, until no more plugins succeed in initialisation or > want to be retried. > > This will fix spider init bugs which is caused in part by its > dependency on Aria for initialisation. > > The reason we need a new return code, instead of treating every > failure as a request for retry, is that it may be impossible to clean > up after a failed plugin initialisation. Take InnoDB for example, it > has a global variable `buf_page_cleaner_is_active`, which may not > satisfy an assertion during a second initialisation try, probably > because InnoDB does not expect the initialisation to be called > twice. A test that may fail because of this is > `encryption.corrupted_during_recovery`, see for example[1], which is > tested at 73835f64b7fc245d38812380685aca03bef72bb5, a previous commit again, commits that you are going to push into the main branche should not refer to commits that you aren't going to push. these are dangling references leading nowhere. > where we retry every failed plugin. > > [1] https://buildbot.mariadb.org/#/builders/369/builds/10107/steps/7/logs/stdio nor references to buildbot logs, they're very short lived, while I can see commits in the server git repo going to 2000. > > diff --git a/sql/sql_plugin.cc b/sql/sql_plugin.cc > index 5a077a934ac..92d74aa51e8 100644 > --- a/sql/sql_plugin.cc > +++ b/sql/sql_plugin.cc > @@ -1435,6 +1435,49 @@ void plugin_unlock_list(THD *thd, plugin_ref *list, size_t count) > DBUG_VOID_RETURN; > } > > +static int plugin_do_initialize(struct st_plugin_int *plugin, uint &state) > +{ > + DBUG_ENTER("plugin_do_initialize"); > + mysql_mutex_assert_not_owner(&LOCK_plugin); > + plugin_type_init init= plugin_type_initialize[plugin->plugin->type]; > + if (!init) > + init= (plugin_type_init) plugin->plugin->init; > + if (init) > + if (int ret= init(plugin)) > + { > + /* Plugin init failed but requested a retry if possible */ > + if (unlikely(ret== HA_ERR_RETRY_INIT)) > + sql_print_warning("Plugin '%s' registration as a %s failed and may " please, don't write "may". It's saying "eh, I don't know, you, user, figure it out". But you should know better than the user. So either say "will be retried" or "won't be retried" (in which case, you can just use the error message below and not mention retry at all). > + "be retried provided it is being loaded with " > + "plugin-load[-add]", > + plugin->name.str, plugin_type_names[plugin->plugin->type].str); > + else > + sql_print_error("Plugin '%s' registration as a %s failed.", > + plugin->name.str, plugin_type_names[plugin->plugin->type].str); > + DBUG_RETURN(ret); > + } > + state= PLUGIN_IS_READY; // plugin->init() succeeded > + > + if (plugin->plugin->status_vars) > + { > + /* > + historical ndb behavior caused MySQL plugins to specify > + status var names in full, with the plugin name prefix. > + this was never fixed in MySQL. > + MariaDB fixes that but supports MySQL style too. > + */ > + SHOW_VAR *show_vars= plugin->plugin->status_vars; > + SHOW_VAR tmp_array[2]= {{plugin->plugin->name, > + (char *) plugin->plugin->status_vars, SHOW_ARRAY}, > + {0, 0, SHOW_UNDEF}}; > + if (strncasecmp(show_vars->name, plugin->name.str, plugin->name.length)) > + show_vars= tmp_array; > + > + if (add_status_vars(show_vars)) > + DBUG_RETURN(1); > + } > + DBUG_RETURN(0); > +} > > static int plugin_initialize(MEM_ROOT *tmp_root, struct st_plugin_int *plugin, > int *argc, char **argv, bool options_only) > @@ -1737,32 +1738,71 @@ int plugin_init(int *argc, char **argv, int flags) > */ > > mysql_mutex_lock(&LOCK_plugin); > + /* List of plugins to reap */ > reap= (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); > *(reap++)= NULL; > + /* List of plugins to retry */ > + retry= (st_plugin_int **) my_alloca((plugin_array.elements+1) * sizeof(void*)); > + *(retry++)= NULL; > > for(;;) > { > + /* Number of plugins that is successfully initialised in a round */ > + int num_initialized; > + do > + { > + num_initialized= 0; > + /* If any plugins failed and requested a retry, clean up before > + retry */ > + while ((plugin_ptr= *(--retry))) > + plugin_ptr->state= PLUGIN_IS_TO_BE_RETRIED; Why do you need a special state for that? You haven't fixed the SHOW PLUGINS command to show it correctly. And I didn't check other conditions, may be some of them need to take the new state into account too, I don't know. Wouldn't it be easier to have a special retry look after the main for() without all its complexity and without a new state. Like (pseudocode, and note that here I've changed the creative trick of walking retry[] backwards until NULL, so now it iterated forward until end): while (retry_start < retry_end) { retry_ptr= retry= retry_start; while ((plugin_ptr= *(retry++))) { mysql_mutex_unlock(&LOCK_plugin); error= plugin_do_initialize(plugin_ptr, state); mysql_mutex_lock(&LOCK_plugin); if (error == HA_ERR_RETRY_INIT) *(retry_ptr++)= plugin_ptr; else if (error) *(reap++)= plugin_ptr; } if (retry == retry_ptr) while (retry_ptr > retry_start) *(reap++)= *(--retry_ptr); retry_end= retry_ptr; } > + retry++; > for (i=0; i < MYSQL_MAX_PLUGIN_TYPE_NUM; i++) > { > HASH *hash= plugin_hash + plugin_type_initialization_order[i]; > for (uint idx= 0; idx < hash->records; idx++) > { > plugin_ptr= (struct st_plugin_int *) my_hash_element(hash, idx); > + if (plugin_ptr->state == PLUGIN_IS_UNINITIALIZED || > + plugin_ptr->state == PLUGIN_IS_TO_BE_RETRIED) > + { > + int error; > if (plugin_ptr->state == PLUGIN_IS_UNINITIALIZED) > { > bool plugin_table_engine= > lex_string_eq(&plugin_table_engine_name, &plugin_ptr->name); > bool opts_only= flags & PLUGIN_INIT_SKIP_INITIALIZATION && > (flags & PLUGIN_INIT_SKIP_PLUGIN_TABLE || > !plugin_table_engine); > - if (plugin_initialize(&tmp_root, plugin_ptr, argc, argv, opts_only)) > + error= plugin_initialize(&tmp_root, plugin_ptr, argc, argv, > + opts_only); > + } else > { > + uint state= plugin_ptr->state; > + mysql_mutex_unlock(&LOCK_plugin); > + error= plugin_do_initialize(plugin_ptr, state); > + mysql_mutex_lock(&LOCK_plugin); > + plugin_ptr->state= state; > + } > + if (error) > + { > plugin_ptr->state= PLUGIN_IS_DYING; > + /* The plugin wants a retry of the initialisation, > + possibly due to dependency on other plugins */ > + if (unlikely(error == HA_ERR_RETRY_INIT)) > + *(retry++)= plugin_ptr; > + else > *(reap++)= plugin_ptr; > } > + else > + num_initialized++; > } > } > } > + /* Only retry if at least one plugin has been initialised > + successfully and at least one has requested a retry during this > + round */ > + } while (num_initialized > 0 && *(retry - 1)); > > /* load and init plugins from the plugin table (unless done already) */ > if (flags & PLUGIN_INIT_SKIP_PLUGIN_TABLE) > @@ -1775,8 +1815,11 @@ int plugin_init(int *argc, char **argv, int flags) > } > > /* > - Check if any plugins have to be reaped > + Merge the retry list to the reap list, then reap the failed > + plugins. Note that during the merge we reverse the order in retry > */ > + while ((plugin_ptr= *(--retry))) > + *(reap++) = plugin_ptr; I already have that in my pseudocode above > while ((plugin_ptr= *(--reap))) > { > mysql_mutex_unlock(&LOCK_plugin); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Numeric datatype with large decimal values
by brajmohan saxena 23 Jun '23

23 Jun '23

Hi Folks, Do we have Or someone already working on creating a Numeric (decimal) data type plugin in MariaDB, Which supports large decimal digits and may be upto infinite. It would be great if someone already has some pointers on it. And let me know the best approach to achieve this type of data type implementation in MariaDB. Thanks in advance. Thanks Braj mohan saxena

1 0

Re: f3aa4d0a215: MDEV-29447 MDEV-26285 MDEV-31338 Refactor spider_db_mbase_util::open_item_func
by Sergei Golubchik 22 Jun '23

22 Jun '23

Hi, Yuchen, On Jun 20, Yuchen Pei wrote: > revision-id: f3aa4d0a215 (mariadb-10.4.30-4-gf3aa4d0a215) > parent(s): f5dceafd0ba > author: Yuchen Pei > committer: Yuchen Pei > timestamp: 2023-06-14 15:02:38 +1000 > message: > > MDEV-29447 MDEV-26285 MDEV-31338 Refactor spider_db_mbase_util::open_item_func > > spider_db_mbase_util::open_item_func() is a monster function. > It is difficult to maintain while it is expected that we need to > modify it when a new SQL function or a new func_type is added. > > We split the function into two distinct functions: one handles the > case of str != NULL and the other handles the case of str == NULL. > > This refactoring was done in a conservative way because we do not > have comprehensive tests on the function. > > It also fixes MDEV-29447 and MDEV-31338 where field items that are > arguments of a func item may be used before created / initialised. > > Note this commit is a port of > 3836098c29ef1b7ff9d5fbde99b690eab73a0df1 (MDEV-26285) to current > versions 10.3+. As I wrote in the first review, please, don't mention transient commit hashes in a permanent commit comment. This commit will, supposedly, be pushed into the main branch. While 3836098c29ef1 will, on the other hand, disappear, as it's not in any of the main branches. You'll have a dangling pointer > Signed-off-by: Yuchen Pei <yuchen.pei(a)mariadb.com> also, it looks strange to sign-off your own commits > > diff --git a/storage/spider/spd_db_mysql.cc b/storage/spider/spd_db_mysql.cc > index a51c1497b02..1ed733105c6 100644 > --- a/storage/spider/spd_db_mysql.cc > +++ b/storage/spider/spd_db_mysql.cc > @@ -5848,94 +5992,12 @@ int spider_db_mbase_util::open_item_func( > { > if (!strncasecmp("utc_timestamp", func_name, func_name_length)) > { > - if (str) > str->length(str->length() - SPIDER_SQL_OPEN_PAREN_LEN); > DBUG_RETURN(spider_db_open_item_string(item_func, NULL, spider, str, > alias, alias_length, dbton_id, use_fields, fields)); > } else if (!strncasecmp("timestampdiff", func_name, func_name_length)) > { > -#ifdef ITEM_FUNC_TIMESTAMPDIFF_ARE_PUBLIC Why did you removed this and not the #else part? > - Item_func_timestamp_diff *item_func_timestamp_diff = > - (Item_func_timestamp_diff *) item_func; > - if (str) > - { > - const char *interval_str; > - uint interval_len; > - switch (item_func_timestamp_diff->int_type) > - { > - case INTERVAL_YEAR: > - interval_str = SPIDER_SQL_YEAR_STR; > - interval_len = SPIDER_SQL_YEAR_LEN; > - break; > - case INTERVAL_QUARTER: > - interval_str = SPIDER_SQL_QUARTER_STR; > - interval_len = SPIDER_SQL_QUARTER_LEN; > - break; > - case INTERVAL_MONTH: > - interval_str = SPIDER_SQL_MONTH_STR; > - interval_len = SPIDER_SQL_MONTH_LEN; > - break; > - case INTERVAL_WEEK: > - interval_str = SPIDER_SQL_WEEK_STR; > - interval_len = SPIDER_SQL_WEEK_LEN; > - break; > - case INTERVAL_DAY: > - interval_str = SPIDER_SQL_DAY_STR; > - interval_len = SPIDER_SQL_DAY_LEN; > - break; > - case INTERVAL_HOUR: > - interval_str = SPIDER_SQL_HOUR_STR; > - interval_len = SPIDER_SQL_HOUR_LEN; > - break; > - case INTERVAL_MINUTE: > - interval_str = SPIDER_SQL_MINUTE_STR; > - interval_len = SPIDER_SQL_MINUTE_LEN; > - break; > - case INTERVAL_SECOND: > - interval_str = SPIDER_SQL_SECOND_STR; > - interval_len = SPIDER_SQL_SECOND_LEN; > - break; > - case INTERVAL_MICROSECOND: > - interval_str = SPIDER_SQL_MICROSECOND_STR; > - interval_len = SPIDER_SQL_MICROSECOND_LEN; > - break; > - default: > - interval_str = ""; > - interval_len = 0; > - break; > - } > - str->length(str->length() - SPIDER_SQL_OPEN_PAREN_LEN); > - if (str->reserve(func_name_length + SPIDER_SQL_OPEN_PAREN_LEN + > - interval_len + SPIDER_SQL_COMMA_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(func_name, func_name_length); > - str->q_append(SPIDER_SQL_OPEN_PAREN_STR, SPIDER_SQL_OPEN_PAREN_LEN); > - str->q_append(interval_str, interval_len); > - str->q_append(SPIDER_SQL_COMMA_STR, SPIDER_SQL_COMMA_LEN); > - } > - if ((error_num = spider_db_print_item_type(item_list[0], NULL, spider, > - str, alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - if (str) > - { > - if (str->reserve(SPIDER_SQL_COMMA_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_COMMA_STR, SPIDER_SQL_COMMA_LEN); > - } > - if ((error_num = spider_db_print_item_type(item_list[1], NULL, spider, > - str, alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - if (str) > - { > - if (str->reserve(SPIDER_SQL_CLOSE_PAREN_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_CLOSE_PAREN_STR, > - SPIDER_SQL_CLOSE_PAREN_LEN); > - } > - DBUG_RETURN(0); > -#else > DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > -#endif > } > } else if (func_name_length == 14) > { Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 3

DBA in Tokyo
by James Tobin 20 Jun '23

20 Jun '23

Hello, I'm working with an employer that is looking to hire a DBA in Tokyo. Consequently, I had hoped some members of this list may like to discuss. I can be reached using "JamesBTobin (at) Gmail (dot) Com". Kind regards, James

1 0

[Maria-developers] Review input for: MDEV-22534 Decorrelate IN subquery
by Sergey Petrunia 01 Jun '23

01 Jun '23

Hi Yuchen, Please find below review input for > commit ba9cf8efeb0b1e1c132d565adb79c98156783f15 > Author: Yuchen Pei <yuchen.pei(a)mariadb.com> > Date: Fri May 19 20:06:31 2023 +1000 > > MDEV-22534 Decorrelate IN subquery > First, general questions: Q1: It seems there is some overlap between Item_in_subselect::exists2in_processor() and Item_exists_subselect::exists2in_processor() Did you consider factoring out common code rather than copying it? (if yes, I'd like to know why it couldn't be done). Q2: Where is the coe that removes the correlated equalities from the subquery's WHERE? (I assume Item_exists_subselect code does it?) > diff --git a/sql/item_subselect.cc b/sql/item_subselect.cc > index 9e6c205ca76..8d893794da3 100644 > --- a/sql/item_subselect.cc > +++ b/sql/item_subselect.cc > @@ -3099,6 +3099,132 @@ static bool find_inner_outer_equalities(Item **conds, > return TRUE; > } > > +/* Checks whether item tree intersects with the free list */ > +static bool intersects_free_list(Item *item, THD *thd) > +{ > + for (const Item *to_find= thd->free_list; to_find; to_find= to_find->next) > + if (item->walk(&Item::find_item_processor, 1, (void *) to_find)) > + return true; > + return false; > +} > + Please try moving this function into a separate file. First name that came to my mind: opt_subq_decorrelate.cc Feel free to suggest better names. The idea is that we should aim for better structure. (This request is only valid if the factoring out suggested above doesn't work out). > +/* De-correlate where conditions in an IN subquery */ This needs a bigger comment describing what the rewrite does. > +bool Item_in_subselect::exists2in_processor(void *opt_arg) > +{ > + THD *thd= (THD *)opt_arg; > + SELECT_LEX *first_select= unit->first_select(); > + JOIN *join= first_select->join; > + bool will_be_correlated; > + Dynamic_array<EQ_FIELD_OUTER> eqs(PSI_INSTRUMENT_MEM, 5, 5); > + List<Item> outer; > + Query_arena *arena= NULL, backup; > + int res= FALSE; > + DBUG_ENTER("Item_in_subselect::exists2in_processor"); > + Please move comments out of the if statement and number the conditions. For examples, see best_access_path(), large if-statement starting with "Don't test table scan if it can't be better." or "EXISTS-to-IN coversion and ORDER BY ... LIMIT clause" in Item_exists_subselect::exists2in_processor(). > + if (!optimizer_flag(thd, OPTIMIZER_SWITCH_EXISTS_TO_IN) || > + /* proceed only if I'm a toplevel IN or a toplevel NOT IN */ > + !(is_top_level_item() || > + (upper_not && upper_not->is_top_level_item())) || what is the reason behind this condition? If we're able to handle top-level NOT IN, why can't we handle subquery in any context? Can we really handle the NOT IN? (I suspect not, it will suffer from the semantics of IN subqueries with NULLs.. I can't find the name for this, it's described e.g. here: https://petrunia.net/2006/11/16/inany-subqueries-null-woes/ ) > + first_select->is_part_of_union() || /* skip if part of a union */ > + first_select->group_list.elements || /* skip if with group by */ > + first_select->with_sum_func || /* skip if aggregation */ > + join->having || /* skip if with having */ > + !join->conds || /* skip if no conds */ > + /* skip if left expr is a single nonscalar subselect */ > + (left_expr->type() == Item::SUBSELECT_ITEM && > + !left_expr->type_handler()->is_scalar_type())) The above seems to be some exotic SQL. Is it something like (select col1,lol2 from t1) IN (select col3,col4 from t2 where ..) and does MariaDB really support this? Please add a comment about this. > + DBUG_RETURN(FALSE); > + /* iterate over conditions, and check whether they can be moved out. */ > + if (find_inner_outer_equalities(&join->conds, eqs)) > + DBUG_RETURN(FALSE); > + /* If we are in the execution of a prepared statement, check for > + intersection with the free list to avoid segfault. Note that the > + check for prepared statement execution is necessary, otherwise it > + will likely always find intersection thus abort the > + transformation. > + > + fixme: this only works when HAVE_PSI_STATEMENT_INTERFACE is > + defined. It causes CI's like amd64-debian-10-debug-embedded to > + fail. Are there other ways to find out we are in the execution of a > + prepared statement? */ I'm looking at the code of activate_stmt_arena_if_needed(). It seems "stmt_arena->is_conventional()" is the check you're looking for. > + if (thd->m_statement_state.m_parent_prepared_stmt) > + { > + for (uint i= 0; i < (uint) eqs.elements(); i++) > + { > + if (intersects_free_list(*eqs.at(i).eq_ref, thd)) > + DBUG_RETURN(FALSE); This check doesn't look good. As far as I understand, it is a temporary measure until related re-execution bug(s) are fixed by Igor (and/or Sanja)? > + } > + } > + /* Determines whether the result will be correlated */ > + { > + List<Item> unused; > + Collect_deps_prm prm= {&unused, // parameters > + unit->first_select()->nest_level_base, // nest_level_base > + 0, // count > + unit->first_select()->nest_level, // nest_level > + FALSE // collect > + }; > + walk(&Item::collect_outer_ref_processor, TRUE, &prm); > + DBUG_ASSERT(prm.count > 0); > + DBUG_ASSERT(prm.count >= (uint)eqs.elements()); > + will_be_correlated= prm.count > (uint)eqs.elements(); See the example pasted in the MDEV: https://jira.mariadb.org/browse/MDEV-22534?focusedCommentId=259705&page=com… Somehow, I get here will_be_correlated==FALSE but Materialization is still not considered for the subquery? > + } > + /* Back up the free list */ > + arena= thd->activate_stmt_arena_if_needed(&backup); > + /* Construct the items for left_expr */ > + if (left_expr->type() == Item::ROW_ITEM) > + for (uint i= 0; i < left_expr->cols(); i++) > + outer.push_back(left_expr->element_index(i)); > + else > + outer.push_back(left_expr); > + const uint offset= first_select->item_list.elements; > + /* Move items to outer and select item list */ > + for (uint i= 0; i < (uint)eqs.elements(); i++) > + { > + Item **eq_ref= eqs.at(i).eq_ref; > + Item_ident *local_field= eqs.at(i).local_field; > + Item *outer_exp= eqs.at(i).outer_exp; > + first_select->item_list.push_back(local_field, thd->mem_root); > + first_select->ref_pointer_array[offset + i]= (Item *)local_field; How do we know that ref_pointer_array has enough room for the new elements? (We can discuss this with Sanja, he's an expert on the topic). I see this check in Item_exists_subselect::exists2in_processor if ((uint)eqs.elements() > (first_select->item_list.elements + first_select->select_n_reserved)) Is it how Item_exists_subselect handles this issue? > + outer.push_back(outer_exp); > + *eq_ref= new (thd->mem_root) Item_int(thd, 1); > + if((*eq_ref)->fix_fields(thd, (Item **)eq_ref)) > + { > + res= TRUE; > + goto out; > + } > + } > + /* Update the left_expr */ > + left_expr= new (thd->mem_root) Item_row(thd, outer); > + if (left_expr->fix_fields(thd, &left_expr)) > + { > + res= TRUE; > + goto out; > + } > + left_expr_orig= left_expr; > + is_correlated= will_be_correlated; > + /* Update any Item_in_optimizer wrapper if exists */ > + if (optimizer) > + { > + optimizer->reset_cache(); > + if (optimizer->fix_left(thd)) > + { > + res= TRUE; > + goto out; > + } > + } > + { > + OPT_TRACE_TRANSFORM(thd, trace_wrapper, trace_transform, > + get_select_lex()->select_number, "IN (SELECT)", > + "decorrelation"); > + } > +out: > + /* Restores the free list */ > + if (arena) > + thd->restore_active_arena(arena, &backup); > + DBUG_RETURN(res); > +} On Mon, Dec 20, 2021 at 09:29:47PM +0300, Sergey Petrunia wrote: > Hello Igor, > > I'm trying examples with this patch: > > create table t1 (a int not null, key(a)); > insert into t1 select seq from seq_1_to_1000; > > First, I debug this example: > > explain select * from t1 where a <10 or a>=10; > > The execution enters key_or(), passes through the new code this patch has added > and key_or() returns NULL. Good. > > But if I modify the example slightly: > > explain select * from t1 where a <20 or a>=10 ; > > Here, key_or() still returns a SEL_ARG object that represents an infinite > range: > > (gdb) fini > Run till exit from #0 key_or (...) > 0x0000555555f1fca2 in tree_or (...) > Value returned is $24 = (SEL_ARG *) 0x7fff70064d88 > (gdb) p dbug_print_sel_arg($24) > $26 = 0x7fff70024390 "SEL_ARG(-inf<=a<=+inf)" > > Should this case be fixed as well? > > On Fri, Dec 17, 2021 at 02:11:40PM -0800, IgorBabaev wrote: > > revision-id: 677643a80986107491b7886441f2828384f0494b (mariadb-10.2.31-1286-g677643a) > > parent(s): 8bb55633699612279744c055e22eeca8d4058273 > > author: Igor Babaev > > committer: Igor Babaev > > timestamp: 2021-12-17 14:11:39 -0800 > > message: > > > > MDEV-27262 Unexpected index intersection with full index scan for an index > > > > If when extracting a range condition foran index from the WHERE condition > > I think the comment wording could be improved also. > > > Range Optimizer sees that the range condition covers the whole index then > > such condition should be discarded because cannot it be used in any range > > scan. In some cases Range Optimizer really does it, but there remained some > > conditions for which it was not done. As a result the optimizer could > > produce index merge plans with the full index scan for one of the indexes > > participating in the index merge. > > This could be observed in one of the test cases from index_merge1.inc > > where a plan with index_merge_sort_union was produced and in the test case > > reported for this bug where a plan with index_merge_sort_intersect was > > produced. In both cases one of two index scans participating in index merge > > ran over the whole index. > > The patch slightly changes the original above mentioned test case from > > index_merge1.inc to be able to produce an intended plan employing > > index_merge_sort_union. The original query was left to show that index > > merge is not used for it anymore. > > It should be noted that for the plan with index_merge_sort_intersect could > > be chosen for execution only due to a defect in the InnoDB code that > > returns wrong estimates for the cardinality of big ranges. > > > > This bug led to serious problems in 10.4+ where the optimization using > > Rowid filters is employed (see mdev-26446). > > > > Approved by Oleksandr Byelkin <sanja(a)mariadb.com> > > > > --- > > mysql-test/include/index_merge1.inc | 8 +++- > > mysql-test/r/index_merge_myisam.result | 12 +++-- > > mysql-test/r/range_innodb.result | 81 ++++++++++++++++++++++++++++++++++ > > mysql-test/t/range_innodb.test | 78 ++++++++++++++++++++++++++++++++ > > sql/opt_range.cc | 7 +++ > > 5 files changed, 181 insertions(+), 5 deletions(-) > > > > diff --git a/mysql-test/include/index_merge1.inc b/mysql-test/include/index_merge1.inc > > index b168a76..440f1f7 100644 > > --- a/mysql-test/include/index_merge1.inc > > +++ b/mysql-test/include/index_merge1.inc > > @@ -150,15 +150,19 @@ explain select * from t0 where > > (((key3 <7 and key7 < 6) or key5 < 2) and (key5 < 5 or key6 < 6)); > > > > explain select * from t0 where > > - ((key3 <5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > + ((key3 < 4 or key5 < 4) and (key1 < 4 or key2 < 4)) > > or > > ((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > > > explain select * from t0 force index(i1, i2, i3, i4, i5, i6 ) where > > - ((key3 <5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > + ((key3 < 4 or key5 < 4) and (key1 < 4 or key2 < 4)) > > or > > ((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > > > +explain select * from t0 force index(i1, i2, i3, i4, i5, i6 ) where > > + ((key3 < 5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > + or > > + ((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > # 8. Verify that "order by" after index merge uses filesort > > select * from t0 where key1 < 5 or key8 < 4 order by key1; > > > > diff --git a/mysql-test/r/index_merge_myisam.result b/mysql-test/r/index_merge_myisam.result > > index 5a23092..a096c34 100644 > > --- a/mysql-test/r/index_merge_myisam.result > > +++ b/mysql-test/r/index_merge_myisam.result > > @@ -173,17 +173,23 @@ or > > id select_type table type possible_keys key key_len ref rows Extra > > 1 SIMPLE t0 index_merge i1,i2,i3,i5,i6,i7 i3,i5 4,4 NULL 11 Using sort_union(i3,i5); Using where > > explain select * from t0 where > > -((key3 <5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > +((key3 < 4 or key5 < 4) and (key1 < 4 or key2 < 4)) > > or > > ((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > id select_type table type possible_keys key key_len ref rows Extra > > 1 SIMPLE t0 ALL i1,i2,i3,i5,i6 NULL NULL NULL 1024 Using where > > explain select * from t0 force index(i1, i2, i3, i4, i5, i6 ) where > > -((key3 <5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > +((key3 < 4 or key5 < 4) and (key1 < 4 or key2 < 4)) > > or > > ((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > id select_type table type possible_keys key key_len ref rows Extra > > -1 SIMPLE t0 index_merge i1,i2,i3,i5,i6 i3,i5 0,4 NULL 1024 Using sort_union(i3,i5); Using where > > +1 SIMPLE t0 index_merge i1,i2,i3,i5,i6 i3,i5 4,4 NULL 1024 Using sort_union(i3,i5); Using where > > +explain select * from t0 force index(i1, i2, i3, i4, i5, i6 ) where > > +((key3 < 5 or key5 < 4) and (key1 < 4 or key2 < 4)) > > +or > > +((key3 >=5 or key5 < 2) and (key5 < 5 or key6 < 6)); > > +id select_type table type possible_keys key key_len ref rows Extra > > +1 SIMPLE t0 ALL i1,i2,i3,i5,i6 NULL NULL NULL 1024 Using where > > select * from t0 where key1 < 5 or key8 < 4 order by key1; > > key1 key2 key3 key4 key5 key6 key7 key8 > > 1 1 1 1 1 1 1 1023 > > diff --git a/mysql-test/r/range_innodb.result b/mysql-test/r/range_innodb.result > > index f2349f2..ccb6da3 100644 > > --- a/mysql-test/r/range_innodb.result > > +++ b/mysql-test/r/range_innodb.result > > @@ -108,3 +108,84 @@ DROP TABLE t0,t1; > > SET @@GLOBAL.debug_dbug = @saved_dbug; > > set @@optimizer_switch= @optimizer_switch_save; > > # End of 10.1 tests > > +# > > +# MDEV-27262: Index intersection with full scan over an index > > +# > > +CREATE TABLE t1 ( > > +id int(10) unsigned NOT NULL AUTO_INCREMENT, > > +p char(32) DEFAULT NULL, > > +es tinyint(3) unsigned NOT NULL DEFAULT 0, > > +er tinyint(3) unsigned NOT NULL DEFAULT 0, > > +x mediumint(8) unsigned NOT NULL DEFAULT 0, > > +PRIMARY KEY (id), > > +INDEX es (es), > > +INDEX x (x), > > +INDEX er (er,x), > > +INDEX p (p) > > +) ENGINE=InnoDB DEFAULT CHARSET=latin1; > > +insert into t1(es,er) select 0, 1 from seq_1_to_45; > > +insert into t1(es,er) select 0, 2 from seq_1_to_49; > > +insert into t1(es,er) select 0, 3 from seq_1_to_951; > > +insert into t1(es,er) select 0, 3 from seq_1_to_1054; > > +insert into t1(es,er) select 0, 6 from seq_1_to_25; > > +insert into t1(es,er) select 0, 11 from seq_1_to_1; > > +insert into t1(es,er) select 1, 1 from seq_1_to_45; > > +insert into t1(es,er) select 1, 2 from seq_1_to_16; > > +insert into t1(es,er) select 1, 3 from seq_1_to_511; > > +insert into t1(es,er) select 1, 4 from seq_1_to_687; > > +insert into t1(es,er) select 1, 6 from seq_1_to_50; > > +insert into t1(es,er) select 1, 7 from seq_1_to_4; > > +insert into t1(es,er) select 1, 11 from seq_1_to_1; > > +insert into t1(es,er) select 2, 1 from seq_1_to_82; > > +insert into t1(es,er) select 2, 2 from seq_1_to_82; > > +insert into t1(es,er) select 2, 3 from seq_1_to_1626; > > +insert into t1(es,er) select 2, 4 from seq_1_to_977; > > +insert into t1(es,er) select 2, 6 from seq_1_to_33; > > +insert into t1(es,er) select 2, 11 from seq_1_to_1; > > +insert into t1(es,er) select 3, 1 from seq_1_to_245; > > +insert into t1(es,er) select 3, 2 from seq_1_to_81; > > +insert into t1(es,er) select 3, 3 from seq_1_to_852; > > +insert into t1(es,er) select 3, 4 from seq_1_to_2243; > > +insert into t1(es,er) select 3, 6 from seq_1_to_44; > > +insert into t1(es,er) select 3, 11 from seq_1_to_1; > > +insert into t1(es,er) select 4, 1 from seq_1_to_91; > > +insert into t1(es,er) select 4, 2 from seq_1_to_83; > > +insert into t1(es,er) select 4, 3 from seq_1_to_297; > > +insert into t1(es,er) select 4, 4 from seq_1_to_2456; > > +insert into t1(es,er) select 4, 6 from seq_1_to_19; > > +insert into t1(es,er) select 4, 11 from seq_1_to_1; > > +update t1 set p='foobar'; > > +update t1 set x=0; > > +set @save_isp=@@innodb_stats_persistent; > > +set global innodb_stats_persistent= 1; > > +analyze table t1; > > +Table Op Msg_type Msg_text > > +test.t1 analyze status OK > > +set optimizer_switch='index_merge_sort_intersection=on'; > > +SELECT * FROM t1 > > +WHERE ((p = 'foo' AND er != 4) OR er = 4 ) AND (es >= 4) LIMIT 2; > > +id p es er x > > +14645 foobar 4 4 0 > > +14646 foobar 4 4 0 > > +EXPLAIN EXTENDED SELECT * FROM t1 > > +WHERE ((p = 'foo' AND er != 4) OR er = 4 ) AND (es >= 4) LIMIT 2; > > +id select_type table type possible_keys key key_len ref rows filtered Extra > > +1 SIMPLE t1 range es,er,p es 1 NULL # 100.00 Using index condition; Using where > > +Warnings: > > +Note 1003 select `test`.`t1`.`id` AS `id`,`test`.`t1`.`p` AS `p`,`test`.`t1`.`es` AS `es`,`test`.`t1`.`er` AS `er`,`test`.`t1`.`x` AS `x` from `test`.`t1` where (`test`.`t1`.`p` = 'foo' and `test`.`t1`.`er` <> 4 or `test`.`t1`.`er` = 4) and `test`.`t1`.`es` >= 4 limit 2 > > +set optimizer_switch='index_merge_sort_intersection=off'; > > +SELECT * FROM t1 > > +WHERE ((p = 'foo' AND er != 4) OR er = 4 ) AND (es >= 4) LIMIT 2; > > +id p es er x > > +14645 foobar 4 4 0 > > +14646 foobar 4 4 0 > > +EXPLAIN EXTENDED SELECT * FROM t1 > > +WHERE ((p = 'foo' AND er != 4) OR er = 4 ) AND (es >= 4) LIMIT 2; > > +id select_type table type possible_keys key key_len ref rows filtered Extra > > +1 SIMPLE t1 range es,er,p es 1 NULL # 100.00 Using index condition; Using where > > +Warnings: > > +Note 1003 select `test`.`t1`.`id` AS `id`,`test`.`t1`.`p` AS `p`,`test`.`t1`.`es` AS `es`,`test`.`t1`.`er` AS `er`,`test`.`t1`.`x` AS `x` from `test`.`t1` where (`test`.`t1`.`p` = 'foo' and `test`.`t1`.`er` <> 4 or `test`.`t1`.`er` = 4) and `test`.`t1`.`es` >= 4 limit 2 > > +set optimizer_switch='index_merge_sort_intersection=default'; > > +set global innodb_stats_persistent= @save_isp; > > +DROP TABLE t1; > > +# End of 10.2 tests > > diff --git a/mysql-test/t/range_innodb.test b/mysql-test/t/range_innodb.test > > index 428e5c2..7420e72 100644 > > --- a/mysql-test/t/range_innodb.test > > +++ b/mysql-test/t/range_innodb.test > > @@ -116,3 +116,81 @@ SET @@GLOBAL.debug_dbug = @saved_dbug; > > set @@optimizer_switch= @optimizer_switch_save; > > > > --echo # End of 10.1 tests > > + > > +--echo # > > +--echo # MDEV-27262: Index intersection with full scan over an index > > +--echo # > > + > > +--source include/have_sequence.inc > > + > > +CREATE TABLE t1 ( > > + id int(10) unsigned NOT NULL AUTO_INCREMENT, > > + p char(32) DEFAULT NULL, > > + es tinyint(3) unsigned NOT NULL DEFAULT 0, > > + er tinyint(3) unsigned NOT NULL DEFAULT 0, > > + x mediumint(8) unsigned NOT NULL DEFAULT 0, > > + PRIMARY KEY (id), > > + INDEX es (es), > > + INDEX x (x), > > + INDEX er (er,x), > > + INDEX p (p) > > +) ENGINE=InnoDB DEFAULT CHARSET=latin1; > > + > > +insert into t1(es,er) select 0, 1 from seq_1_to_45; > > +insert into t1(es,er) select 0, 2 from seq_1_to_49; > > +insert into t1(es,er) select 0, 3 from seq_1_to_951; > > +insert into t1(es,er) select 0, 3 from seq_1_to_1054; > > +insert into t1(es,er) select 0, 6 from seq_1_to_25; > > +insert into t1(es,er) select 0, 11 from seq_1_to_1; > > +insert into t1(es,er) select 1, 1 from seq_1_to_45; > > +insert into t1(es,er) select 1, 2 from seq_1_to_16; > > +insert into t1(es,er) select 1, 3 from seq_1_to_511; > > +insert into t1(es,er) select 1, 4 from seq_1_to_687; > > +insert into t1(es,er) select 1, 6 from seq_1_to_50; > > +insert into t1(es,er) select 1, 7 from seq_1_to_4; > > +insert into t1(es,er) select 1, 11 from seq_1_to_1; > > +insert into t1(es,er) select 2, 1 from seq_1_to_82; > > +insert into t1(es,er) select 2, 2 from seq_1_to_82; > > +insert into t1(es,er) select 2, 3 from seq_1_to_1626; > > +insert into t1(es,er) select 2, 4 from seq_1_to_977; > > +insert into t1(es,er) select 2, 6 from seq_1_to_33; > > +insert into t1(es,er) select 2, 11 from seq_1_to_1; > > +insert into t1(es,er) select 3, 1 from seq_1_to_245; > > +insert into t1(es,er) select 3, 2 from seq_1_to_81; > > +insert into t1(es,er) select 3, 3 from seq_1_to_852; > > +insert into t1(es,er) select 3, 4 from seq_1_to_2243; > > +insert into t1(es,er) select 3, 6 from seq_1_to_44; > > +insert into t1(es,er) select 3, 11 from seq_1_to_1; > > +insert into t1(es,er) select 4, 1 from seq_1_to_91; > > +insert into t1(es,er) select 4, 2 from seq_1_to_83; > > +insert into t1(es,er) select 4, 3 from seq_1_to_297; > > +insert into t1(es,er) select 4, 4 from seq_1_to_2456; > > +insert into t1(es,er) select 4, 6 from seq_1_to_19; > > +insert into t1(es,er) select 4, 11 from seq_1_to_1; > > +update t1 set p='foobar'; > > +update t1 set x=0; > > +set @save_isp=@@innodb_stats_persistent; > > +set global innodb_stats_persistent= 1; > > +analyze table t1; > > + > > +let $q= > > +SELECT * FROM t1 > > + WHERE ((p = 'foo' AND er != 4) OR er = 4 ) AND (es >= 4) LIMIT 2; > > + > > +set optimizer_switch='index_merge_sort_intersection=on'; > > +eval $q; > > +--replace_column 9 # > > +eval EXPLAIN EXTENDED $q; > > + > > +set optimizer_switch='index_merge_sort_intersection=off'; > > +# execution of $q and explain for it led to an assertion failure in 10.4 > > +# (with the optimizer switch rowid_filter set to 'on') > > +eval $q; > > +--replace_column 9 # > > +eval EXPLAIN EXTENDED $q; > > +set optimizer_switch='index_merge_sort_intersection=default'; > > + > > +set global innodb_stats_persistent= @save_isp; > > +DROP TABLE t1; > > + > > +--echo # End of 10.2 tests > > diff --git a/sql/opt_range.cc b/sql/opt_range.cc > > index f3f1843..2e05b88 100644 > > --- a/sql/opt_range.cc > > +++ b/sql/opt_range.cc > > @@ -9413,6 +9413,13 @@ key_or(RANGE_OPT_PARAM *param, SEL_ARG *key1,SEL_ARG *key2) > > key2->copy_min(tmp); > > if (!(key1=key1->tree_delete(tmp))) > > { // Only one key in tree > > + if (key2->min_flag & NO_MIN_RANGE && > > + key2->max_flag & NO_MAX_RANGE) > > + { > > + if (key2->maybe_flag) > > + return new SEL_ARG(SEL_ARG::MAYBE_KEY); > > + return 0; // Always true OR > > + } > > key1=key2; > > key1->make_root(); > > key2=key2_next; > > BR > Sergei > -- > Sergei Petrunia, Software Developer > MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net > > -- BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net

2 3

Re: [Maria-developers] c457f237511: MDEV-30984 Online ALTER table is denied with non-informative error messages
by Sergei Golubchik 29 May '23

29 May '23

Hi, Nikita, comments below refer to a combined diff of both commits On May 23, Nikita Malyavin wrote: > revision-id: c457f237511 (mariadb-11.0.1-124-gc457f237511) > parent(s): f4b04ec534d > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-15 21:07:50 +0300 > message: > > MDEV-30984 Online ALTER table is denied with non-informative error messages Two main thoughts. First, users will generally don't do show warnings after an error - mariadb already puts some useful info into a warning after an error in some (few) cases, and it's always a surprise when I write in a bug report "use SHOW WARNINGS". It's not very intuitive. And in your case it's even unnecessary, because your error message is LOCK=NONE is not supported. Reason: %s. Try LOCK=SHARED and "Reason" part is perfectly suited for an explanation that you generate, instead of a misleading "COPY algorithm requires a lock" (because it doesn't). Second thought, your messages are hard-coded and won't be translated even if the user will configure the server to use, say, spanish or Japanese error messages. See how it was solved for "COPY algorithm requires a lock", ER_ALTER_OPERATION_NOT_SUPPORTED_REASON_COPY. But I think the message will still be comprehensible if you'll simply print the offending sql clause as the reason, this won't need any localization. See below where I've shown what your error messages would look like if you do that. If you think that's too cryptic, then, please, use the ER_ALTER_OPERATION_NOT_SUPPORTED_REASON_COPY approach. Two more comments at the very end but I'll repeat them here, because they're easy to overlook: AUTO_INCREMENT instead of AUTOINC, and a test for DROP SYSTEM VERSIONING. > diff --git a/mysql-test/main/alter_table_online.result b/mysql-test/main/alter_table_online.result > index 8b726ccb591..bcf7afea9df 100644 > --- a/mysql-test/main/alter_table_online.result > +++ b/mysql-test/main/alter_table_online.result > @@ -4,6 +4,10 @@ > create table t (a int); > alter ignore table t add primary key (a), algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: ALTER IGNORE. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 ALTER IGNORE TABLE is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > drop table t; > # > # MDEV-28771 Assertion `table->in_use&&tdc->flushed' failed after ALTER > @@ -92,12 +96,20 @@ on update cascade) engine=InnoDB; > insert into t2 values (1),(2),(3); > alter table t2 add c int, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: FOREIGN KEY ... ON UPDATE CASCADE. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Tables with CASCADE/SET NULL foreign keys are incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t2 add c int, algorithm=inplace, lock=none; > create or replace table t2 (b int, foreign key (b) > references t1 (a) > on delete set null) engine=InnoDB; > alter table t2 add c int, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: FOREIGN KEY ... ON DELETE SET NULL. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Tables with CASCADE/SET NULL foreign keys are incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t2 add c int, algorithm=inplace, lock=none; > create or replace table t2 (b int, foreign key (b) > references t1 (a) > @@ -117,6 +129,10 @@ b int references t1 (b) on update cascade) engine=InnoDB; > insert into t2 values (1, 1),(2, 2); > alter table t2 add c int, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: FOREIGN KEY ... ON DELETE SET NULL. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Tables with CASCADE/SET NULL foreign keys are incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t2 add c int, algorithm=copy; > alter table t2 add d int, algorithm=inplace; > drop table t2, t1; > @@ -132,6 +148,10 @@ period for system_time (row_start, row_end)) > engine=innodb with system versioning; > alter table t1 add c int, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: BIGINT GENERATED ALWAYS AS ROW_START. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Transaction-versioned tables are incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t1 add c int, algorithm=inplace; > alter table t1 add d int, lock=none; > set system_versioning_alter_history= default; > @@ -143,6 +163,10 @@ drop table t1; > create table t (a serial, b int) engine=innodb; > alter table t drop a, modify b serial, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... SERIAL. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > set statement sql_mode= NO_AUTO_VALUE_ON_ZERO for > alter table t drop a, modify b serial, algorithm=copy, lock=none; > create or replace table t (a serial, b int) engine=innodb; > @@ -157,6 +181,10 @@ t CREATE TABLE `t` ( > # Only unsafe approach is fine because of possible collisions. > alter table t modify a int, modify b serial, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... SERIAL. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > # > # Check that we treat autoinc columns correctly modify old autoinc is > # fine, adding new autoinc for existed column is unsafe. > @@ -180,22 +208,38 @@ t CREATE TABLE `t` ( > ) ENGINE=InnoDB DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci > alter table t drop b, change c c serial, algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... SERIAL. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > # Check existed unique keys. > create or replace table t(a int, b int not null, c int not null, d int); > # No unique in the old table; > alter table t add unique(b, c), modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t add unique(a, b); > # Unique in the old table has nulls; > alter table t modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > alter table t add unique(b, c); > # Change unique'scolumn; > alter table t change b x int, modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > ERROR 0A000: LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED ERROR 0A000: LOCK=NONE is not supported. Reason: MODIFY COLUMN ... AUTO_INCREMENT. Try LOCK=SHARED > +show warnings; > +Level Code Message > +Note 1846 Adding AUTOINC to an existing column for a table without a primary key is incompatible with LOCK=NONE, ALGORITHM=COPY > +Error 1846 LOCK=NONE is not supported. Reason: COPY algorithm requires a lock. Try LOCK=SHARED > # Finally good. > alter table t modify d int auto_increment, add key(d), > algorithm=copy, lock=none; > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > index 99570008ddb..22447db42af 100644 > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -9916,6 +9916,67 @@ bool online_alter_check_autoinc(const THD *thd, const Alter_info *alter_info, > return online; > } > > +bool online_alter_is_supported(THD *thd, const Alter_info *alter_info, > + const TABLE *table) > +{ > + const char *reason= NULL; > + List<FOREIGN_KEY_INFO> fk_list; > + > + if (alter_info->flags & ALTER_DROP_SYSTEM_VERSIONING) > + { > + reason= "DROP SYSTEM VERSIONING is"; no test for that > + goto unsupported; > + } > + > + if (thd->lex->ignore) > + { > + reason= "ALTER IGNORE TABLE is"; > + goto unsupported; > + } > + > + if (table->s->tmp_table) > + { > + reason= "Temporary tables are"; > + goto unsupported; > + } > + > + if (table->versioned(VERS_TRX_ID)) > + { > + reason= "Transaction-versioned tables are"; > + goto unsupported; > + } > + > + table->file->get_foreign_key_list(thd, &fk_list); > + for (auto &fk: fk_list) > + { > + if (fk_modifies_child(fk.delete_method) > + || fk_modifies_child(fk.update_method)) > + { > + reason= "Tables with CASCADE/SET NULL foreign keys are"; > + goto unsupported; > + } > + } > + > + if (!online_alter_check_autoinc(thd, alter_info, table)) > + { > + reason= "Adding AUTOINC to an existing column for a table without a " AUTO_INCREMENT, please, don't abbrev user visible messages > + "primary key is"; > + goto unsupported; > + } > + > + return true; > + > +unsupported: > + if (alter_info->requested_lock == Alter_info::ALTER_TABLE_LOCK_NONE > + && alter_info->algorithm(thd) == Alter_info::ALTER_TABLE_ALGORITHM_COPY) > + push_warning_printf(thd, Sql_condition::WARN_LEVEL_NOTE, > + ER_ALTER_OPERATION_NOT_SUPPORTED_REASON, > + "%s incompatible with " > + "LOCK=NONE, ALGORITHM=COPY", > + reason); > + return false; > +} > + > > /** > Alter table Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 4

Re: [Maria-developers] 56e2cd20ed3: MDEV-31033 ER_KEY_NOT_FOUND upon online COPY ALTER on a partitioned table
by Sergei Golubchik 29 May '23

29 May '23

Hi, Nikita, On May 24, Nikita Malyavin wrote: > revision-id: 56e2cd20ed3 (mariadb-11.0.1-114-g56e2cd20ed3) > parent(s): 47f5c7ae7e4 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-10 01:49:42 +0300 > message: > > MDEV-31033 ER_KEY_NOT_FOUND upon online COPY ALTER on a partitioned table > > The row events were applied "twice": once for the ha_partition, and one > more time for the underlying storage engine. > > There's no such problem in binlog/rpl, because ha_partiton::row_logging > is normally set to false. > > The fix makes the events replicate only when the handler is a root handler. > We will try to *guess* this by comparing it to table->file. The same > approach is used in the MDEV-21540 fix, 231feabd. The assumption is made, > that the row methods are only called for table->file (and never for a > cloned handler), hence the assertions are added in ha_innobase and > ha_myisam to make sure that this is true at least for those engines > > Also closes MDEV-31040, however the test is not included, since we have no > convenient way to construct a deterministic version. > > diff --git a/sql/handler.cc b/sql/handler.cc > index 3fcbb5171d6..d8aca5bc80b 100644 > --- a/sql/handler.cc > +++ b/sql/handler.cc > @@ -7280,7 +7280,7 @@ int handler::binlog_log_row(const uchar *before_record, > log_func, row_logging_has_trans); > > #ifdef HAVE_REPLICATION > - if (unlikely(!error && table->s->online_alter_binlog)) > + if (unlikely(!error && table->s->online_alter_binlog && is_root_handler())) > error= binlog_log_row_online_alter(table, before_record, after_record, > log_func); > #endif // HAVE_REPLICATION > @@ -7800,7 +7800,7 @@ int handler::ha_write_row(const uchar *buf) > DBUG_RETURN(error); > > /* > - NOTE: this != table->file is true in 3 cases: > + NOTE: is_root_handler() is false in 3 cases: may be move this comment closer to is_root_handler() ? > > 1. under copy_partitions() (REORGANIZE PARTITION): that does not > require long unique check as it does not introduce new rows or new index. > @@ -7810,7 +7810,7 @@ int handler::ha_write_row(const uchar *buf) > 3. under ha_mroonga::wrapper_write_row() > */ > > - if (table->s->long_unique_table && this == table->file) > + if (table->s->long_unique_table && is_root_handler()) > { > DBUG_ASSERT(inited == NONE || lookup_handler != this); > if ((error= check_duplicate_long_entries(buf))) > @@ -7862,13 +7862,13 @@ int handler::ha_update_row(const uchar *old_data, const uchar *new_data) > error= ha_check_overlaps(old_data, new_data); > > /* > - NOTE: this != table->file is true under partition's ha_update_row(): > + NOTE: is_root_handler() is false under partition's ha_update_row(): > check_duplicate_long_entries_update() was already done by > ha_partition::ha_update_row(), no need to check it again for each single > partition. Same applies to ha_mroonga wrapper. > */ > > - if (!error && table->s->long_unique_table && this == table->file) > + if (!error && table->s->long_unique_table && is_root_handler()) > error= check_duplicate_long_entries_update(new_data); > table->status= saved_status; > > @@ -7913,6 +7913,12 @@ int handler::ha_update_row(const uchar *old_data, const uchar *new_data) > return error; > } > > + > +bool handler::is_root_handler() const > +{ > + return this == table->file; > +} better make it inline in handler.h > + > /* > Update first row. Only used by sequence tables > */ > diff --git a/sql/handler.h b/sql/handler.h > index fb0eba83c38..9b903e7ebce 100644 > --- a/sql/handler.h > +++ b/sql/handler.h > @@ -3502,6 +3502,10 @@ class handler :public Sql_alloc > return extra(HA_EXTRA_NO_KEYREAD); > } > > +protected: > + bool is_root_handler() const; > + > +public: > int check_collation_compatibility(); > int check_long_hash_compatibility() const; > int ha_check_for_upgrade(HA_CHECK_OPT *check_opt); > diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc > index f035437b1d7..94d9da3486b 100644 > --- a/storage/innobase/handler/ha_innodb.cc > +++ b/storage/innobase/handler/ha_innodb.cc > @@ -7750,6 +7750,8 @@ ha_innobase::write_row( > > DBUG_ENTER("ha_innobase::write_row"); > > + DBUG_ASSERT(is_root_handler() || table->file->ht != innodb_hton_ptr); instead of putting this assert into every handler's implementation and in many methods, why not to have it in handler ha_ methods or may be even in the is_root_handler itself? You can compare this->ht with table->file->ht. For example, like DBUG_ASSERT(this == table->file || this->ht != table->file->ht); > + > trx_t* trx = thd_to_trx(m_user_thd); > > /* Validation checks before we commence write_row operation. */ > @@ -8479,6 +8481,8 @@ ha_innobase::update_row( > > DBUG_ENTER("ha_innobase::update_row"); > > + DBUG_ASSERT(is_root_handler() || table->file->ht != innodb_hton_ptr); > + > if (is_read_only()) { > DBUG_RETURN(HA_ERR_TABLE_READONLY); > } else if (!trx_is_started(trx)) { > @@ -8653,6 +8657,8 @@ ha_innobase::delete_row( > > DBUG_ENTER("ha_innobase::delete_row"); > > + DBUG_ASSERT(is_root_handler() || table->file->ht != innodb_hton_ptr); > + > if (is_read_only()) { > DBUG_RETURN(HA_ERR_TABLE_READONLY); > } else if (!trx_is_started(trx)) { > diff --git a/storage/myisam/ha_myisam.cc b/storage/myisam/ha_myisam.cc > index 4057c3ea859..b7fadd86a23 100644 > --- a/storage/myisam/ha_myisam.cc > +++ b/storage/myisam/ha_myisam.cc > @@ -983,6 +983,7 @@ int ha_myisam::close(void) > > int ha_myisam::write_row(const uchar *buf) > { > + DBUG_ASSERT(is_root_handler() || table->file->ht != ht); > /* > If we have an auto_increment column and we are writing a changed row > or a new row, then update the auto_increment value in the record. > @@ -1957,11 +1958,13 @@ bool ha_myisam::is_crashed() const > > int ha_myisam::update_row(const uchar *old_data, const uchar *new_data) > { > + DBUG_ASSERT(is_root_handler() || table->file->ht != ht); > return mi_update(file,old_data,new_data); > } > > int ha_myisam::delete_row(const uchar *buf) > { > + DBUG_ASSERT(is_root_handler() || table->file->ht != ht); > return mi_delete(file,buf); > } > > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 5

[Maria-developers] MTR fails on undefined {cpus} array in My/SysInfo.pm on armhf and armel
by Otto Kekäläinen 28 May '23

28 May '23

Hi! I noticed that the autopkgtests on Debian on armhf and armel that run the mariadb-test-run fail on error: starting mysql-test-tun.pl... Logging: ./mysql-test-run.pl --force --testcase-timeout=120 --suite-timeout=540 --retry=3 -- ... Collecting tests... Installing system database... Can't use an undefined value as an ARRAY reference at lib/My/SysInfo.pm line 166. Logs for full context: https://ci.debian.net/data/autopkgtest/unstable/armel/m/mariadb/33870767/lo… https://ci.debian.net/data/autopkgtest/unstable/armhf/m/mariadb/33870768/lo… This line 166 in mysql-test/lib/My/SysInfo.pm has: # Return the number of cpus found sub num_cpus { my ($self)= @_; return int(@{$self->{cpus}}) or confess "INTERNAL ERROR: No cpus in list"; } Can somebody with strong Perl skills help me understand 1) Where this cpus array is inherited from? 2) Why might it have now regressed in MariaDB 10.11.3 and why is this erroring only on armel/armhf but not other architectures? 3) What should I do wrap this section in so that if cpus is not set, it would just default to '1'? - Otto

3 6

Re: [Maria-developers] c75f774d278: MDEV-31058 ER_KEY_NOT_FOUND upon concurrent CHANGE column autoinc and DML
by Sergei Golubchik 27 May '23

27 May '23

Hi, Nikita, On May 24, Nikita Malyavin wrote: > revision-id: c75f774d278 (mariadb-11.0.1-118-gc75f774d278) > parent(s): 319207175bd > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-15 19:43:49 +0300 > message: > > MDEV-31058 ER_KEY_NOT_FOUND upon concurrent CHANGE column autoinc and DML > > When column is changed to autoinc, ALTER TABLE may update zero/NULL values, > if NO_AUTO_VALUE_ON_ZERO mode is not enabled. > > Forbid this for LOCK=NONE for the unreliable cases. > The cases are described in online_alter_check_autoinc. just disable online for any ALTER that adds an autoinc attribute to an existing column. See MDEV. > diff --git a/mysql-test/main/alter_table_online.test b/mysql-test/main/alter_table_online.test > index b9e147c812d..7ebe553aabb 100644 > --- a/mysql-test/main/alter_table_online.test > +++ b/mysql-test/main/alter_table_online.test > @@ -135,3 +135,71 @@ alter table t1 add d int, lock=none; > set system_versioning_alter_history= default; > drop table t1; > > +--echo # > +--echo # MDEV-31058 ER_KEY_NOT_FOUND upon concurrent CHANGE column autoinc > +--echo # and DML > +--echo # > +create table t (a serial, b int) engine=innodb; > +--error ER_ALTER_OPERATION_NOT_SUPPORTED_REASON > +alter table t drop a, modify b serial, algorithm=copy, lock=none; > + > +set statement sql_mode= NO_AUTO_VALUE_ON_ZERO for > +alter table t drop a, modify b serial, algorithm=copy, lock=none; I don't think it should work, see my comment in the MDEV > +--echo # Finally good. > +alter table t modify d int auto_increment, add key(d), > + algorithm=copy, lock=none; no, it's not good either, as far as I understand > + > +drop table t; lots of unvisible spaces at line ends, please remove them > + > +--echo # MDEV-31172 Server crash or ASAN errors in online_alter_check_autoinc > +create table t (a int, b int, c char(8), key(a,b,c)); > +alter table t modify c int auto_increment key, algorithm=copy; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

[Maria-developers] Regression in crossbuilding MariaDB
by Otto Kekäläinen 26 May '23

26 May '23

Hello! While working on https://salsa.debian.org/mariadb-team/mariadb-server/-/merge_requests/46 (Import new upstream version 10.11.3 to Debian) I noticed that amd64->arm64 crossbuilding no longer works. The CI shows that Salsa-CI job `test-crossbuild-arm64` no longer passes. Based on Salsa-CI runs I can see that it [still worked in upstream commit e55e761e](https://salsa.debian.org/otto/mariadb-server/-/jobs/4217220) but [failed in upstream commit 06b443be](https://salsa.debian.org/otto/mariadb-server/-/jobs/4217664), thus some of these commits in between regressed it: ``` 06b443be34e Nes CC 3.3 4a668c18926 MDEV-29401 InnoDB history list length increased in 10.6 compared to 10.5 f272463b028 Cleanup of MDEV-14974: --port ignored for --host=localhost bb1d1dc8465 Merge 10.5 into 10.6 c5e50c48bba MDEV-30837: Remove usage of AWK in autobake-debs.sh f99a891858a MDEV-30837: Remove usage of AWK from Debian init and postinst scripts 5740638c4c3 MDEV-31132 Deadlock between DDL and purge of InnoDB history d4265fbde58 MDEV-26055: Correct the formula for adaptive flushing 898320b5f8a MDEV-30804 addendum for 10.6+ branches c22ab93f8af MDEV-26827 fixup: Prevent a hang in LRU eviction 818d5e48146 Merge 10.5 into 10.6 0976afec889 MDEV-31114 Assertion !...is_waiting() failed in os_aio_wait_until_no_pending_writes() 2c567b2fa37 MDEV-30996 insert.. select in presence of full text index freezes all other commits at commit time 51e62cb3b3b MDEV-26782 InnoDB temporary tablespace: reclaiming of free space does not work 204e7225dce Cleanup: MONITOR_EXISTING trx_undo_slots_used, trx_undo_slots_cached 86767bcc0f1 MDEV-29593 Purge misses a chance to free not-yet-reused undo pages 40eff3f8681 MDEV-26827 fixup: hangs and !os_aio_pending_writes() assertion failures ``` Example of error in (cross)build log: ``` Building CXX object CMakeFiles/cmTC_9b8d6.dir/src.cxx.o /usr/lib/ccache/c++ -DHAVE_INCONSISTENT_MISSING_DESTRUCTOR_OVERRIDE -g -O2 -ffile-prefix-map=/builds/otto/mariadb-server/debian/output/source_dir=. -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -Wdate-time -D_FORTIFY_SOURCE=2 -pie -fPIC -fstack-protector --param=ssp-buffer-size=4 -Wall -Wextra -Woverloaded-virtual -Wconversion -g -Wsuggest-override -Winconsistent-missing-destructor-override -o CMakeFiles/cmTC_9b8d6.dir/src.cxx.o -c /builds/otto/mariadb-server/debian/output/source_dir/builddir-native/CMakeFiles/CMakeScratch/TryCompile-Nn98w6/src.cxx c++: error: unrecognized command-line option '-Winconsistent-missing-destructor-override' gmake[3]: *** [CMakeFiles/cmTC_9b8d6.dir/build.make:78: CMakeFiles/cmTC_9b8d6.dir/src.cxx.o] Error 1 gmake[3]: Leaving directory '/builds/otto/mariadb-server/debian/output/source_dir/builddir-native/CMakeFiles/CMakeScratch/TryCompile-Nn98w6' gmake[2]: *** [Makefile:127: cmTC_9b8d6/fast] Error 2 ``` Is this enough information for somebody to guess what regressed? - Otto

4 6

[Maria-developers] Understanding binlog group commit (MDEV-232, MDEV-532, MDEV-25611, MDEV-18959)
by Marko Mäkelä 22 May '23

22 May '23

Hi Kristian, I am happy to see you being active on the maria-developers list. I don’t think we have collaborated in the past. I have been working on the InnoDB storage engine under MySQL 4.0 through 5.7 (and some 8.0) since 2003, and since 2016 under MariaDB Server, mostly 10.1 and later. I understand that you were an early contributor to MariaDB and implemented some durability I/O optimization between the "distributed transaction" of binlog and the InnoDB write-ahead log (redo log, ib_logfile0). I think that some improvement in this area is long overdue. I find it hard to read the high-level descriptions. Nowhere in https://jira.mariadb.org/browse/MDEV-232 or https://jira.mariadb.org/browse/MDEV-532 I see any mention of any key terminology: * InnoDB log sequence number (LSN): a logical point of time for any persistent InnoDB changes. Measured in bytes. Mini-transactions (atomic changes of InnoDB pages) start and end at some LSN. Corresponds to the Oracle system change number (SCN). * Binlog offset: a logical point of time for any changes at the replication level. Measured in bytes. Replication events such as statements or row-level events start and end at some binlog offset. * The LSN or binlog offset of a COMMIT, XA PREPARE, XA ROLLBACK, or XA COMMIT. There is quite a bit of code in InnoDB that I would like to get rid of. Some of it is related to the data members trx_t::flush_log_later and trx_t::must_flush_log_later. There is also the confusingly named handlerton::commit_checkpoint_request that has nothing to do with log checkpoints, but with some kind of a durability request that fails to identify the minimum logical time up to which everything is desired to be durable. (My talk https://fosdem.org/2022/schedule/event/mariadb_innodb/ explains what a log checkpoint is and how recovery works.) Some years ago, in MDEV-21534 Vladislav Vaintroub introduced write_lock and flush_lock in InnoDB. This works pretty well: a Sysbench workload with innodb_flush_log_at_trx_commit=1 will linearly improve throughput as you add more client connection threads. I think that something similar would be beneficial for the binlog, to keep track of the latest written offset and the latest write offset after which fdatasync() has completed. I am not familiar with the binlog code. A colleague recently claimed to me that LOCK_commit_ordered is a huge bottleneck. As you can see, the write_lock and flush_lock are better because the lock requests include a notion of logical time: If everything up to "my" time has been written or durably written, I do not care to wait. In MDEV-27774 (MariaDB Server 10.8) this was improved further: multiple mini-transactions can copy their local log buffer to the shared log_sys.buf concurrently. It seems to me that the missing tracking of logical time in the data structures is causing hangs like MDEV-25611. I feel that the design is based on a wrong assumption that the binlog-covered transactions are the only source of InnoDB log writes. That has never been the case: For example, there have always been background threads or tasks that purge the history of committed transactions. The task MDEV-18959 (ensuring that one fdatasync() per durable transaction suffices) is also related to this. I think that the simplest solution (assuming that moving the binlog to a bunch of transactional tables is out of the question) is to ensure that one of the logs (binlog and other write-ahead logs) is always ahead. If we choose that the binlog must be ahead, then in InnoDB the write_lock acquisition would have to be preceded by a binlog_write_lock acquisition that will ensure that any relevant changes of persistent state (COMMIT, XA PREPARE, XA ROLLBACK, XA COMMIT) that could be part of the InnoDB log write must have been durably written to the binlog. With best regards, Marko -- Marko Mäkelä, Lead Developer InnoDB MariaDB plc

3 7

Re: [Maria-developers] Understanding binlog group commit (MDEV-232, MDEV-532, MDEV-25611, MDEV-18959)
by Marko Mäkelä 16 May '23

16 May '23

On Tue, May 16, 2023 at 1:37 PM Andrei Elkin <andrei.elkin(a)mariadb.com> wrote: > Can you comment on chances for `flush_lsn` never advance to the value of > `write_lsn` when/because Innodb has nothing to commit (so at such a time > RM (or PURGE or FLUSH LOGS that have been mentioned too) might be run)? If the last write to the log was an InnoDB internal one, it could be done by log_write_up_to(lsn, false). The log_sys.flushed_to_disk_lsn would only be advanced (to somewhere closer to log_sys.lsn) if there was a subsequent call to log_write_up_to(lsn, true). Log checkpoint would advance log_sys.flushed_to_disk_lsn, as should a user transaction commit. This could be a possible explanation. -- Marko Mäkelä, Lead Developer InnoDB MariaDB plc

1 0

Re: [Maria-developers] d67c3f88883: MDEV-27744 InnoDB: Failing assertion: !cursor->index->is_committed() in row0ins.cc (from row_ins_sec_index_entry_by_modify) | Assertion `0' failed in row_upd_sec_index_entry (debug) | Corruption
by Sergei Golubchik 11 May '23

11 May '23

Hi, Alexander, A question: if the frm will *always* be created in the canonical, non-ORACLE mode and always parsed in the same mode, will we even need these schema-qualified native functions? okay, the schema-qualified trick will help to avoid creating REGEX_REPLACE_ORACLE and so on. But is it needed to fix MDEV-27744? More comments below On Apr 27, Alexander Barkov wrote: > revision-id: d67c3f88883 (mariadb-10.3.33-61-gd67c3f88883) > parent(s): 7355f7b1f5c > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2022-04-22 15:35:16 +0400 > message: > > MDEV-27744 InnoDB: Failing assertion: !cursor->index->is_committed() in row0ins.cc (from row_ins_sec_index_entry_by_modify) | Assertion `0' failed in row_upd_sec_index_entry (debug) | Corruption > > The crash happened with an indexed virtual column whose > value is evaluated using a function that has a different meaning > in sql_mode='' vs sql_mode=ORACLE: ... good comment > diff --git a/sql/sql_partition.cc b/sql/sql_partition.cc > index 59d2ea60715..8e83ee02782 100644 > --- a/sql/sql_partition.cc > +++ b/sql/sql_partition.cc > @@ -2591,11 +2591,9 @@ char *generate_partition_syntax_for_frm(THD *thd, partition_info *part_info, > HA_CREATE_INFO *create_info, > Alter_info *alter_info) > { > - sql_mode_t old_mode= thd->variables.sql_mode; > - thd->variables.sql_mode &= ~MODE_ANSI_QUOTES; > + Sql_mode_save_for_frm_handling sql_mode_save(thd); right. thanks > char *res= generate_partition_syntax(thd, part_info, buf_length, > true, create_info, alter_info); > - thd->variables.sql_mode= old_mode; > return res; > } > > diff --git a/sql/unireg.cc b/sql/unireg.cc > index efbb5e773b4..3b37944afa5 100644 > --- a/sql/unireg.cc > +++ b/sql/unireg.cc > @@ -652,6 +649,7 @@ static bool pack_expression(String *buf, Virtual_column_info *vcol, > static bool pack_vcols(String *buf, List<Create_field> &create_fields, > List<Virtual_column_info> *check_constraint_list) > { > + Sql_mode_save_for_frm_handling sql_mode_save(current_thd); well, it was done above pack_vcols(), because the caller had access to the thd. Why not to keep it that way? > List_iterator<Create_field> it(create_fields); > Create_field *field; > > diff --git a/sql/item_cmpfunc.cc b/sql/item_cmpfunc.cc > index f41414f8ae9..0f5c7519d07 100644 > --- a/sql/item_cmpfunc.cc > +++ b/sql/item_cmpfunc.cc > @@ -3292,9 +3292,24 @@ void Item_func_case_simple::print(String *str, enum_query_type query_type) > } > > > +Item_func_decode_oracle * > +Item_func_decode_oracle::create(THD *thd, const LEX_CSTRING &name, > + List<Item> *item_list) > +{ > + if (unlikely(!item_list || item_list->elements < 3)) > + { > + wrong_param_count_error(oracle_schema_ref.name(), name); > + return NULL; > + } > + return new (thd->mem_root) Item_func_decode_oracle(thd, *item_list); > +} why? items are normally created in the Create_xxx::create* methods, defined in item_create.cc. Why did you delegate that to static methods in Item_func_xxx ? > + > + > void Item_func_decode_oracle::print(String *str, enum_query_type query_type) > { > - str->append(func_name()); > + print_sql_mode_dependent_name(str, query_type, > + oracle_schema_ref, > + Item_func_decode_oracle::func_name()); > str->append('('); > args[0]->print(str, query_type); > for (uint i= 1, count= when_count() ; i <= count; i++) > diff --git a/sql/mysqld.h b/sql/mysqld.h > index d40e1d170d0..54069c3cd5a 100644 > --- a/sql/mysqld.h > +++ b/sql/mysqld.h > @@ -741,12 +741,25 @@ enum enum_query_type > QT_ITEM_SUBSELECT_ID_ONLY, > > QT_SHOW_SELECT_NUMBER= (1<<10), > + > + //QT_ITEM_IDENT_DISABLE_DB_TABLE_NAMES= (1 <<11), -- this is taken in 10.5 > + > + /// Print sql_mode-dependent functions with the schema qualifier > + /// even if the currently implied (by sql_mode) schema is equal to > + /// to the function schema, e.g. mariadb_schema.concat('a'). > + QT_ITEM_FUNC_FORCE_SCHEMA_NAME= (1 << 12), > + > + /// Print for FRM file. Focus on parse-back. > + /// e.g. VIEW expressions and virtual column expressions > + QT_FOR_FRM= (1 << 13), > + > /// This is used for EXPLAIN EXTENDED extra warnings / Be more detailed > /// Be more detailed than QT_EXPLAIN. > /// Perhaps we should eventually include QT_ITEM_IDENT_SKIP_CURRENT_DATABASE > /// here, as it would give better readable results > QT_EXPLAIN_EXTENDED= QT_TO_SYSTEM_CHARSET| > - QT_SHOW_SELECT_NUMBER, > + QT_SHOW_SELECT_NUMBER| > + QT_ITEM_FUNC_FORCE_SCHEMA_NAME, don't do this. let's try to keep the behavior *exactly as before* for people who don't switch between sql_modes all the time on the second thought. why do you need QT_ITEM_FUNC_FORCE_SCHEMA_NAME at all? Besides EXPLAIN EXTENDED (where we'd better not use it), it's only used in generating view's sql for I_S. It's not supposed to be used for generating views, so why would it need QT_ITEM_FUNC_FORCE_SCHEMA_NAME? May be it can be deleted altogether? > > // If an expression is constant, print the expression, not the value > // it evaluates to. Should be used for error messages, so that they > diff --git a/sql/sql_class.h b/sql/sql_class.h > index 8268ebe0bcd..6b988938f0b 100644 > --- a/sql/sql_class.h > +++ b/sql/sql_class.h > @@ -6589,6 +6589,50 @@ class Sql_mode_save > sql_mode_t old_mode; // SQL mode saved at construction time. > }; > > + > +/* > + Save the current sql_mode. Switch off sql_mode flags which can prevent > + normal parsing of VIEWs, expressions in generated columns. > + Restore the old sql_mode on destructor. > +*/ > +class Sql_mode_save_for_frm_handling: public Sql_mode_save > +{ > +public: > + Sql_mode_save_for_frm_handling(THD *thd) > + :Sql_mode_save(thd) > + { > + /* > + - MODE_REAL_AS_FLOAT affect only CREATE TABLE parsing > + + MODE_PIPES_AS_CONCAT affect expression parsing > + + MODE_ANSI_QUOTES affect expression parsing > + + MODE_IGNORE_SPACE affect expression parsing > + - MODE_IGNORE_BAD_TABLE_OPTIONS affect only CREATE/ALTER TABLE parsing > + * MODE_ONLY_FULL_GROUP_BY affect execution > + * MODE_NO_UNSIGNED_SUBTRACTION affect execution > + - MODE_NO_DIR_IN_CREATE affect table creation only > + - MODE_POSTGRESQL compounded from other modes > + - MODE_ORACLE affects Item creation (e.g for CONCAT) should be '+' for MODE_ORACLE, not '-', right? > + - MODE_MSSQL compounded from other modes > + - MODE_DB2 compounded from other modes > + - MODE_MAXDB affect only CREATE TABLE parsing > + - MODE_NO_KEY_OPTIONS affect only SHOW > + - MODE_NO_TABLE_OPTIONS affect only SHOW > + - MODE_NO_FIELD_OPTIONS affect only SHOW > + - MODE_MYSQL323 affect only SHOW > + - MODE_MYSQL40 affect only SHOW > + - MODE_ANSI compounded from other modes > + (+ transaction mode) > + ? MODE_NO_AUTO_VALUE_ON_ZERO affect UPDATEs > + + MODE_NO_BACKSLASH_ESCAPES affect expression parsing > + + MODE_EMPTY_STRING_IS_NULL affect expression parsing > + */ > + thd->variables.sql_mode&= ~(MODE_PIPES_AS_CONCAT | MODE_ANSI_QUOTES | > + MODE_IGNORE_SPACE | MODE_NO_BACKSLASH_ESCAPES | > + MODE_ORACLE | MODE_EMPTY_STRING_IS_NULL); > + }; > +}; > + > + > class Abort_on_warning_instant_set > { > THD *m_thd; > diff --git a/sql/sql_schema.h b/sql/sql_schema.h > index 7c8f284d526..36914e36990 100644 > --- a/sql/sql_schema.h > +++ b/sql/sql_schema.h > @@ -33,6 +36,34 @@ class Schema > { > return src; > } > + > + > + /** > + Find the native function builder associated with a given function name. > + @param thd The current thread > + @param name The native function name > + @return The native function builder associated with the name, or NULL > + */ > + virtual Create_func *find_native_function_builder(THD *thd, > + const LEX_CSTRING &name) > + const; > + /** > + Find a native function builder, return an error if not found, > + build an Item otherwise. > + */ > + Item *make_item_func_call_native(THD *thd, > + const Lex_ident_sys &name, > + List<Item> *args) const; > + > + // Builders for native SQL function with a special syntax in sql_yacc.yy > + virtual Item *make_item_func_substr(THD *thd, > + const Lex_substring_spec_st &spec) const; > + > + virtual Item *make_item_func_trim(THD *thd, const Lex_trim_st &spec) const; > + virtual Item *make_item_func_replace(THD *thd, > + Item *subj, > + Item *find, > + Item *replace) const; Why no Lex_replace_spec_st ? > /* > For now we have *hard-coded* compatibility schemas: > schema_mariadb, schema_oracle, schema_maxdb. > @@ -66,5 +97,6 @@ class Schema > > > extern Schema mariadb_schema; > +extern const Schema &oracle_schema_ref; ??? why did you declare mariadb_schema extern, but oracle_schema_ref only via a const ref? > > #endif // SQL_SCHEMA_H_INCLUDED > diff --git a/sql/item_create.h b/sql/item_create.h > index 894e9777b8d..3fadaecb090 100644 > --- a/sql/item_create.h > +++ b/sql/item_create.h > @@ -215,9 +205,48 @@ struct Native_func_registry > Create_func *builder; > }; > > + > +class Native_functions_hash: public HASH > +{ > +public: > + Native_functions_hash() > + { > + bzero(this, sizeof(*this)); > + } > + ~Native_functions_hash() > + { > + /* > + No automatic free. > + The the upper level code should call cleanup() explicitly. why is that? > + */ > + DBUG_ASSERT(!records); > + } > + bool init(size_t count); > + bool init(const Native_func_registry array[], size_t count) > + { > + return init(count) || append(array); > + } > + bool append(const Native_func_registry array[]); > + bool remove(const Native_func_registry array[]); > + void cleanup(); > + /** > + Find the native function builder associated with a given function name. > + @param thd The current thread > + @param name The native function name > + @return The native function builder associated with the name, or NULL > + */ > + Create_func *find(THD *thd, const LEX_CSTRING &name) const; > +}; > + > +extern Native_functions_hash native_functions_hash; > +extern Native_functions_hash native_functions_hash_oracle_overrides; > + > +extern const Native_func_registry func_array[]; > +extern const size_t func_array_length; > + > int item_create_init(); > -int item_create_append(Native_func_registry array[]); > -int item_create_remove(Native_func_registry array[]); > +int item_create_append(const Native_func_registry array[]); > +int item_create_remove(const Native_func_registry array[]); > void item_create_cleanup(); > > Item *create_func_dyncol_create(THD *thd, List<DYNCALL_CREATE_DEF> &list); > diff --git a/sql/sql_schema.cc b/sql/sql_schema.cc > index 0bf4a63c2f8..e2993f51c6d 100644 > --- a/sql/sql_schema.cc > +++ b/sql/sql_schema.cc > @@ -78,3 +89,86 @@ Schema *Schema::find_implied(THD *thd) > return &maxdb_schema; > return &mariadb_schema; > } > + > + > +Create_func * > +Schema::find_native_function_builder(THD *thd, const LEX_CSTRING &name) const > +{ > + return native_functions_hash.find(thd, name); > +} > + > + > +Create_func * > +Schema_oracle::find_native_function_builder(THD *thd, > + const LEX_CSTRING &name) const > +{ > + Create_func *create= native_functions_hash_oracle_overrides.find(thd, name); > + if (create) > + return create; > + return native_functions_hash.find(thd, name); may be populate native_functions_hash_oracle will all names to do only one lookup here? this will be used often in oracle mode > +} > + > + > +Item *Schema::make_item_func_call_native(THD *thd, > + const Lex_ident_sys &name, > + List<Item> *args) const > +{ > + Create_func *builder= find_native_function_builder(thd, name); > + if (builder) > + return builder->create_func(thd, &name, args); > + my_error(ER_FUNCTION_NOT_DEFINED, MYF(0), name.str); > + return NULL; > +} > + > + > +Item *Schema::make_item_func_trim(THD *thd, const Lex_trim_st &spec) const > +{ > + return spec.make_item_func_trim_std(thd); why does trim is constructed via helper in spec, but substr and replace are not? > +} > + > + > +Item *Schema::make_item_func_substr(THD *thd, > + const Lex_substring_spec_st &spec) const > +{ > + return spec.m_for ? > + new (thd->mem_root) Item_func_substr(thd, spec.m_subject, spec.m_from, > + spec.m_for) : > + new (thd->mem_root) Item_func_substr(thd, spec.m_subject, spec.m_from); > +} > + > + > +Item *Schema_oracle::make_item_func_substr(THD *thd, > + const Lex_substring_spec_st &spec) const > +{ > + return spec.m_for ? > + new (thd->mem_root) Item_func_substr_oracle(thd, spec.m_subject, > + spec.m_from, > + spec.m_for) : > + new (thd->mem_root) Item_func_substr_oracle(thd, spec.m_subject, > + spec.m_from); > +} > + > + > +Item *Schema_oracle::make_item_func_trim(THD *thd, > + const Lex_trim_st &spec) const > +{ > + return spec.make_item_func_trim_oracle(thd); > +} > + > + > +Item *Schema::make_item_func_replace(THD *thd, > + Item *subj, > + Item *find, > + Item *replace) const > +{ > + return new (thd->mem_root) Item_func_replace(thd, subj, find, replace); > +} > + > + > +Item *Schema_oracle::make_item_func_replace(THD *thd, > + Item *subj, > + Item *find, > + Item *replace) const > +{ > + return new (thd->mem_root) Item_func_replace_oracle(thd, subj, find, replace); > +} > diff --git a/sql/item_func.h b/sql/item_func.h > index 754b1cd1eb2..52bf7f661d1 100644 > --- a/sql/item_func.h > +++ b/sql/item_func.h > @@ -56,8 +56,117 @@ class Item_func :public Item_func_or_sum > bool check_argument_types_can_return_text(uint start, uint end) const; > bool check_argument_types_can_return_date(uint start, uint end) const; > bool check_argument_types_can_return_time(uint start, uint end) const; > + > + void print_schema_qualified_name(String *to, > + const LEX_CSTRING &schema_name, > + const char *function_name) const > + { > + // e.g. oracle_schema.func() > + to->append(schema_name); > + to->append('.'); > + to->append(function_name); > + } > + > + void print_name_with_optional_suffix(String *to, > + const char *function_name, > + const LEX_CSTRING &suffix) const > + { > + // e.g. func_oracle() > + to->append(function_name); > + if (suffix.length) > + { > + to->append("_"); > + to->append(suffix); > + } > + } > + > + void print_sql_mode_dependent_name(String *to, enum_query_type query_type, > + const Schema &schema, > + const char *function_name, > + const LEX_CSTRING &suffix) const > + { > + if (query_type & QT_ITEM_FUNC_FORCE_SCHEMA_NAME) > + { > + DBUG_ASSERT((query_type & QT_FOR_FRM) == 0); > + print_schema_qualified_name(to, schema.name(), function_name); > + } > + else if (query_type & QT_FOR_FRM) > + { > + // e.g. substr_oracle() > + DBUG_ASSERT((query_type & QT_ITEM_FUNC_FORCE_SCHEMA_NAME) == 0); > + print_name_with_optional_suffix(to, function_name, suffix); > + } > + else if (&schema != Schema::find_implied(current_thd)) > + print_schema_qualified_name(to, schema.name(), function_name); > + else > + to->append(function_name); > + } > + > + void print_sql_mode_dependent_name(String *to, enum_query_type query_type, > + const Schema &schema, > + const char *function_name) const > + { > + static const LEX_CSTRING oracle= {STRING_WITH_LEN("oracle")}; can it be oracle_schema.name() instead? > + static const LEX_CSTRING empty= {NULL, 0}; > + const LEX_CSTRING suffix= &schema == &oracle_schema_ref ? oracle : empty; > + print_sql_mode_dependent_name(to, query_type, > + schema, function_name, suffix); > + } > + void print_sql_mode_dependent_name(String *to, enum_query_type query_type, > + const char *function_name) const > + { > + print_sql_mode_dependent_name(to, query_type, *schema(), function_name); > + } > + void print_sql_mode_dependent(String *to, enum_query_type query_type) > + { > + print_sql_mode_dependent_name(to, query_type, func_name()); > + print_args_parenthesized(to, query_type); > + } > public: > > + // Print an error message for a builtin-schema qualified function call > + static void wrong_param_count_error(const LEX_CSTRING &schema_name, > + const LEX_CSTRING &func_name); > + > + // Check that the number of arguments is greater or equal to "expected" > + static bool create_check_args_ge(const LEX_CSTRING &name, > + const List<Item> *item_list, > + uint expected) > + { > + DBUG_ASSERT(expected > 0); > + if (!item_list || item_list->elements < expected) > + { > + my_error(ER_WRONG_PARAMCOUNT_TO_NATIVE_FCT, MYF(0), name.str); > + return true; > + } > + return false; > + } > + > + // Check that the number of arguments is less or equal to "expected" > + static bool create_check_args_le(const LEX_CSTRING &name, > + const List<Item> *item_list, > + uint expected) > + { > + DBUG_ASSERT(expected > 0); > + if (!item_list || item_list->elements > expected) > + { > + my_error(ER_WRONG_PARAMCOUNT_TO_NATIVE_FCT, MYF(0), name.str); > + return true; > + } > + return false; > + } > + > + // Check that the number of arguments is in the allowed range > + static bool create_check_args_between(const LEX_CSTRING &name, > + const List<Item> *item_list, > + uint min_arg_count, > + uint max_arg_count) > + { > + DBUG_ASSERT(min_arg_count < max_arg_count); > + return create_check_args_ge(name, item_list, min_arg_count) || > + create_check_args_le(name, item_list, max_arg_count); > + } > + > table_map not_null_tables_cache; > > enum Functype { UNKNOWN_FUNC,EQ_FUNC,EQUAL_FUNC,NE_FUNC,LT_FUNC,LE_FUNC, > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy > index 3770488f3f4..e0bafc75a36 100644 > --- a/sql/sql_yacc.yy > +++ b/sql/sql_yacc.yy > @@ -11235,6 +11236,34 @@ function_call_generic: > if (unlikely(!($$= Lex->make_item_func_call_generic(thd, &$1, &$3, &$5, $7)))) > MYSQL_YYABORT; > } > + | ident_cli '.' REPLACE '(' expr ',' expr ',' expr ')' > + { > + if (unlikely(!($$= Lex->make_item_func_replace(thd, $1, $3, > + $5, $7, $9)))) > + MYSQL_YYABORT; > + } > + | ident_cli '.' SUBSTRING '(' substring_operands ')' > + { > + if (unlikely(!($$= Lex->make_item_func_substr(thd, $1, $3, $5)))) > + MYSQL_YYABORT; > + } > + | ident_cli '.' TRIM '(' trim_operands ')' > + { > + if (unlikely(!($$= $5.make_item_func_trim(thd, $1, $3)))) > + MYSQL_YYABORT; > + } > + /* > + We don't add a qualified syntax for TRIM_ORACLE here, > + as this syntax is not absolutely required: > + SELECT mariadb_schema.TRIM_ORACLE(..); > + What absolutely required is only: > + SELECT mariadb_schema.TRIM(..); > + Adding a qualified syntax for TRIM_ORACLE would be tricky because > + it is a non-reserved keyword. To avoid new shift/reduce conflicts > + it would require grammar changes, like introducing a new rule > + ident_step2_cli (which would include everything that ident_cli > + includes but TRIM_ORACLE). Did you add a qualified syntax for any other xxx_ORACLE functions? > + */ > ; > > fulltext_options: > diff --git a/sql/sql_lex.cc b/sql/sql_lex.cc > index 7f1b362d91d..6b0a351f76e 100644 > --- a/sql/sql_lex.cc > +++ b/sql/sql_lex.cc > @@ -2008,7 +2023,59 @@ bool Lex_input_stream::get_7bit_or_8bit_ident(THD *thd, uchar *last_char) > } > > > +/* > + Resolve special SQL functions that have a qualified syntax in sql_yacc.yy. > + These functions are not listed in the native function registry > + because of a special syntax, or a reserved keyword: > + > + mariadb_schema.SUBSTRING('a' FROM 1 FOR 2) -- Special syntax > + mariadb_schema.TRIM(BOTH ' ' FROM 'a') -- Special syntax > + mariadb_schema.REPLACE('a','b','c') -- Verb keyword > +*/ > + > +int Lex_input_stream::find_keyword_qualified_special_func(Lex_ident_cli_st *str, > + uint length) const > +{ > + /* > + There are many other special functions, see the following grammar rules: > + function_call_keyword > + function_call_nonkeyword > + Here we resolve only those that have a qualified syntax to handle > + different behavior in different @@sql_mode settings. > + > + Other special functions do not work in qualified context: > + SELECT mariadb_schema.year(now()); -- Function year is not defined > + SELECT mariadb_schema.now(); -- Function now is not defined > + > + We don't resolve TRIM_ORACLE here, because it does not have > + a qualified syntax yet. Search for "trim_operands" in sql_yacc.yy > + to find more comments. > + */ > + static LEX_CSTRING funcs[]= > + { > + {STRING_WITH_LEN("SUBSTRING")}, > + {STRING_WITH_LEN("SUBSTR")}, > + {STRING_WITH_LEN("TRIM")}, > + {STRING_WITH_LEN("REPLACE")} > + }; > + > + int tokval= find_keyword(str, length, true); > + if (!tokval) > + return 0; > + for (size_t i= 0; i < array_elements(funcs); i++) > + { > + CHARSET_INFO *cs= system_charset_info; > + if (!cs->coll->strnncollsp(cs, may be strcasecmp? or with cs=latin1? you know the repertoire is ascii, so you can cut corners here. > + (const uchar *) m_tok_start, length, > + (const uchar *) funcs[i].str, funcs[i].length)) > + return tokval; > + } > + return 0; > +} > + > + > -int Lex_input_stream::scan_ident_sysvar(THD *thd, Lex_ident_cli_st *str) > +int Lex_input_stream::scan_ident_common(THD *thd, Lex_ident_cli_st *str, > + Ident_mode mode) > { > uchar last_char; > uint length; > @@ -8035,30 +8133,49 @@ bool LEX::add_grant_command(THD *thd, enum_sql_command sql_command_arg, > } > > > +const Schema * > +LEX::find_func_schema_by_name_or_error(const Lex_ident_sys &schema, > + const Lex_ident_sys &func) > +{ > + Schema *res= Schema::find_by_name(schema); > + if (res) > + return res; > + Database_qualified_name qname(schema, func); > + my_error(ER_FUNCTION_NOT_DEFINED, MYF(0), ErrConvDQName(&qname).ptr()); eh, not quite. I can define a function with the same name as a native function in any db. E.g. create test.replace(a int) returns int and it'll be callable as test.replace(1) So ER_FUNCTION_NOT_DEFINED is wrong here. But note that even though test.substr() is possible, it should not allow substr() special syntax. > + return NULL; > +} > -Item *LEX::make_item_func_substr(THD *thd, Item *a, Item *b, Item *c) > -{ > - return (thd->variables.sql_mode & MODE_ORACLE) ? > - new (thd->mem_root) Item_func_substr_oracle(thd, a, b, c) : > - new (thd->mem_root) Item_func_substr(thd, a, b, c); > -} > > > -Item *LEX::make_item_func_substr(THD *thd, Item *a, Item *b) > -{ > - return (thd->variables.sql_mode & MODE_ORACLE) ? > - new (thd->mem_root) Item_func_substr_oracle(thd, a, b) : > - new (thd->mem_root) Item_func_substr(thd, a, b); > -} > +Item *LEX::make_item_func_substr(THD *thd, > + const Lex_ident_cli_st &schema_name_cli, > + const Lex_ident_cli_st &func_name_cli, > + const Lex_substring_spec_st &spec) > +{ > + Lex_ident_sys schema_name(thd, &schema_name_cli); > + Lex_ident_sys func_name(thd, &func_name_cli); > + if (schema_name.is_null() || func_name.is_null()) > + return NULL; // EOM > + const Schema *schema= find_func_schema_by_name_or_error(schema_name, > + func_name); > + return schema ? schema->make_item_func_substr(thd, spec) : NULL; > +} > > > Item *LEX::make_item_func_replace(THD *thd, > + const Lex_ident_cli_st &schema_name_cli, > + const Lex_ident_cli_st &func_name_cli, > Item *org, > Item *find, > Item *replace) > { > - return (thd->variables.sql_mode & MODE_ORACLE) ? > - new (thd->mem_root) Item_func_replace_oracle(thd, org, find, replace) : > - new (thd->mem_root) Item_func_replace(thd, org, find, replace); > + Lex_ident_sys schema_name(thd, &schema_name_cli); > + Lex_ident_sys func_name(thd, &func_name_cli); > + if (schema_name.is_null() || func_name.is_null()) > + return NULL; // EOM > + const Schema *schema= find_func_schema_by_name_or_error(schema_name, > + func_name); > + return schema ? schema->make_item_func_replace(thd, org, find, replace) : > + NULL; > } > > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 2

[Maria-developers] START ALTER replication test cases
by Kristian Nielsen 10 May '23

10 May '23

Hi Sachin, Andrei, I'm looking at test cases for START ALTER / Lag-free Alter on Slave. And I happened to notice something in test case rpl.rpl_start_alter_bugs: --connection master set global binlog_alter_two_phase=true; # ... ALTER TABLE t1 DROP PRIMARY KEY; ALTER TABLE t1 ADD UNIQUE KEY ui (i); ALTER TABLE t1 ADD PRIMARY KEY (i); Note the set *global* binlog_alter_two_phase. This will not affect the testcase at all, as the ALTER TABLE queries are using the *session* value (which defaults to false). This can be verified by checking the binlog after running the testcase, mysql-test/var/mysqld.1/data/master-bin.000001 . It has only normal ALTER TABLE query events, no START ALTER / COMMIT ALTER. So it looks strange that a start_alter test doesn't use START ALTER, and I just wanted to point this out to you in case it is a mistake and maybe the test coverage of START ALTER is less than intended? A quick grep shows similar SET GLOBAL in other start_alter tests, but I did not investigate if they have the same issue. Hope this helps, - Kristian.

2 2

Re: [Maria-developers] da5a72f32d4: MDEV-31033 ER_KEY_NOT_FOUND upon online COPY ALTER on a partitioned table
by Sergei Golubchik 10 May '23

10 May '23

Hi, Nikita, On May 05, Nikita Malyavin wrote: > revision-id: da5a72f32d4 (mariadb-11.0.1-114-gda5a72f32d4) > parent(s): 0d6584c019c > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-01 01:15:41 +0300 > message: > > MDEV-31033 ER_KEY_NOT_FOUND upon online COPY ALTER on a partitioned table > > The row events were applied "twice": once for the ha_partition, and > one more time for the underlying storage engine. > > There's no such problem in binlog/rpl, because > ha_partiton::row_logging is normally set to false. s/ha_partiton/ha_partition/ > The proposed fix adds table->skip_online_logging field, preventing the > underlying engines from replicating online events. Only upper layer > engine should do this. > > table->skip_online_logging is initialized to 0 together with the table > itself. Normally it is always 0, except for the intermediate > overriding part, once it is checked in handler::binlog_log_row. > > Note on MDEV-31040 (Server crashes in MYSQL_LOG::is_open upon ALTER on > partitioned table): > First the solution was to override table->s->online_alter_binlog, and > set it to NULL during the partitioning operation, which was > ridiculously wrong: no-one is supposed to modify that field outside > ALTER! That field write access is protected by EXCLUSIVE metadata > lock. I don't think you need to list your different (non-working) attemps to fix the bug in the commit comment. Just explain the fix that works, not the fix that doesn't. > In comparison, table->skip_online_logging is a connection-local field. > > The test is a non-deterministic one. It wasn't convenient to construct > a deterministic test here, besides for the case that will never be > repeated. Is it still the case? Your test in alter_table_online_debug.test with debug_sync - is it non-deterministic? > diff --git a/sql/handler.cc b/sql/handler.cc > index 3fcbb5171d6..a7e229e8f8d 100644 > --- a/sql/handler.cc > +++ b/sql/handler.cc > @@ -7280,7 +7280,8 @@ int handler::binlog_log_row(const uchar *before_record, > log_func, row_logging_has_trans); > > #ifdef HAVE_REPLICATION > - if (unlikely(!error && table->s->online_alter_binlog)) > + if (unlikely(!error && table->s->online_alter_binlog && > + !table->skip_online_logging)) this doesn't need any juggling with skip_online_logging. This problem is already solved for, say, long uniques. You can use this == table->file condition to filter out individual partitions and have your binlog_log_row_online_alter called only for the main ha_partition handler. > error= binlog_log_row_online_alter(table, before_record, after_record, > log_func); > #endif // HAVE_REPLICATION Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 5

Re: [Maria-developers] 24c653be25a: unpack_row: unpack a correct number of fields
by Sergei Golubchik 06 May '23

06 May '23

Hi, Nikita, On May 05, Nikita Malyavin wrote: > revision-id: 24c653be25a (mariadb-11.0.1-123-g24c653be25a) > parent(s): c1cbda5a2c8 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-04 20:49:43 +0300 > message: > > unpack_row: unpack a correct number of fields > > diff --git a/sql/rpl_record.cc b/sql/rpl_record.cc > index 908d2c39d97..30bd80bdfc7 100644 > --- a/sql/rpl_record.cc > +++ b/sql/rpl_record.cc > @@ -228,24 +228,30 @@ int unpack_row(const rpl_group_info *rgi, TABLE *table, uint const colcnt, > const TABLE *conv_table= rpl_data.conv_table; > DBUG_PRINT("debug", ("Table data: tabldef: %p, conv_table: %p", > tabledef, conv_table)); > - > + bool is_online_alter= rpl_data.is_online_alter(); > DBUG_ASSERT(rgi); > > - for (field_ptr= begin_ptr; field_ptr < end_ptr && *field_ptr; ++field_ptr) > + for (field_ptr= begin_ptr; field_ptr < end_ptr > + /* In Online Alter conv_table can be wider than > + original table, but we need to unpack it all. */ > + && (*field_ptr || is_online_alter); I failed to understand this, could you please explain? The comment didn't help much, unfortunately. The test case didn't either. > + ++field_ptr) > { > + intptr fidx= field_ptr - begin_ptr; intptr is a variant of int with sizeof(intptr) == sizeof(void*). it's needed to cast a pointer to int, do some math, cast back. here you need my_ptrdiff_t or ptrdiff_t > /* > If there is a conversion table, we pick up the field pointer to > the conversion table. If the conversion table or the field > pointer is NULL, no conversions are necessary. > */ > - Field *conv_field= conv_table ? conv_table->field[field_ptr - begin_ptr] : NULL; > + Field *conv_field= conv_table ? conv_table->field[fidx] : NULL; > Field *const f= conv_field ? conv_field : *field_ptr; > +#ifdef DBUG_TRACE > + Field *dbg= is_online_alter ? f : *field_ptr; > +#endif > DBUG_PRINT("debug", ("Conversion %srequired for field '%s' (#%ld)", > conv_field ? "" : "not ", > - (*field_ptr)->field_name.str, > - (long) (field_ptr - begin_ptr))); > + dbg->field_name.str, > + (long) ())); eh? you surely meant `(long) fidx` ? how did it even compile? > DBUG_ASSERT(f != NULL); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

Re: [Maria-developers] c5ce597f06a: MDEV-31043 ER_KEY_NOT_FOUND upon concurrent ALTER and transaction
by Sergei Golubchik 06 May '23

06 May '23

Hi, Nikita, On May 03, Nikita Malyavin wrote: > revision-id: c5ce597f06a (mariadb-11.0.1-115-gc5ce597f06a) > parent(s): da5a72f32d4 > author: Nikita Malyavin > committer: Nikita Malyavin > timestamp: 2023-05-01 01:15:41 +0300 > message: > > MDEV-31043 ER_KEY_NOT_FOUND upon concurrent ALTER and transaction > > Non-transactional engines changes are visible immediately the row operation > succeeds, so we should apply the changes to the online buffer immediately > after the statement ends. > > diff --git a/sql/log.cc b/sql/log.cc > index f30ce3bcbd1..1d98455c2b3 100644 > --- a/sql/log.cc > +++ b/sql/log.cc > @@ -7701,23 +7701,35 @@ static int binlog_online_alter_end_trans(THD *thd, bool all, bool commit) > auto *binlog= cache.sink_log; > DBUG_ASSERT(binlog); > > - bool do_commit= commit || !cache.hton->rollback || > - cache.hton->flags & HTON_NO_ROLLBACK; // Aria > + bool do_commit= (commit && is_ending_transaction) > + || cache.hton->flags & HTON_NO_ROLLBACK // Aria > + || !cache.hton->rollback; > + > + error= binlog_flush_pending_rows_event(thd, false, true, binlog, &cache); you've lost the useful comment about STMT_END. and you flush events for rollback too that seems like a waste. I personally would've kept the old structure of if()'s. Like - bool do_commit= commit || !cache.hton->rollback || + bool non_trans= !cache.hton->rollback || and - if (do_commit) + if (commit || non_trans) with - if (is_ending_transaction) + if (is_ending_transaction || non_trans) That would've called binlog flush only where it needs to be. > + > if (do_commit) > { > - // do not set STMT_END for last event to leave table open in altering thd > - error= binlog_flush_pending_rows_event(thd, false, true, binlog, &cache); > - if (is_ending_transaction) > + /* > + If the cache wasn't reinited to write, then it remains empty after > + the last write. > + */ > + if (cache.cache_log.type != READ_CACHE && !error) this is a confusing new condition. are you trying to avoid locking a mutex for an empty cache? If yes, you can check my_b_bytes_in_cache(), that'd be more clear. > { > mysql_mutex_lock(binlog->get_log_lock()); > error= binlog->write_cache(thd, &cache.cache_log); > mysql_mutex_unlock(binlog->get_log_lock()); > } > - else > - cache.store_prev_position(); > } > - else if (!is_ending_transaction) > + else if (!commit) // rollback > + { > cache.restore_prev_position(); > + } > + else add // trans engine, end of statement > + { > + DBUG_ASSERT(!is_ending_transaction); > + cache.store_prev_position(); > + } Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 3

Re: [Maria-developers] d25439f1cbb: MDEV-31140: FLUSH BINARY LOGS DELETE_DOMAIN_ID=(D) can errorneously delete active domains
by Kristian Nielsen 02 May '23

02 May '23

Andrei Elkin <andrei.elkin(a)mariadb.com> writes: >> Want to review a small patch for https://jira.mariadb.org/browse/MDEV-31140 ? > Thanks for a very good piece of analysis and the refinement! > I agree with both. Great, thanks for review! Do you have an opinion into which branch I should push it? The fix should be low-risk. I'm thinking maybe 11.0 or 10.11 is fine, since the bug does not affect normal operation. But I'm open to other suggestions, whatever is the current policy for this kind of bug? > The existed test was more than a bit unlucky, to confess, > to let the problem stay unnoticed. Yes, there is a good test coverage, so a bit unlucky, agree. - Kristian. > On the occasion of the International workers' day, let me wish you more > contribution, and inspiration for how to make not just this server > better alone :-)!!! Thanks, and all the best to you too :-) - Kristian. >> Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: >> >>> revision-id: d25439f1cbb79e5467b4249792130bfe524e3f10 (mariadb-10.11.2-21-gd25439f1cbb) >>> parent(s): 656c2e18b1e9ea5d0314745f3988d126eedbc22a >>> author: Kristian Nielsen >>> committer: Kristian Nielsen >>> timestamp: 2023-04-27 12:16:07 +0200 >>> message: >>> >>> MDEV-31140: FLUSH BINARY LOGS DELETE_DOMAIN_ID=(D) can errorneously delete active domains >>> >>> Fix the code in rpl_binlog_state::drop_domain(), so that _all_ entries for >>> the domain in the binlog state must match an entry in the initial GTID_LIST, >>> not just one entry match. >>> >>> Signed-off-by: Kristian Nielsen <knielsen(a)knielsen-hq.org> >>> >>> --- >>> .../r/binlog_flush_binlogs_delete_domain.result | 14 ++++++++++--- >>> .../t/binlog_flush_binlogs_delete_domain.test | 13 +++++++++++- >>> sql/rpl_gtid.cc | 24 ++++++++++++---------- >>> 3 files changed, 36 insertions(+), 15 deletions(-) >>> >>> diff --git a/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result b/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result >>> index fdcfb4bfa01..1c11191802f 100644 >>> --- a/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result >>> +++ b/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result >>> @@ -46,15 +46,23 @@ Warning 1076 The current gtid binlog state is incompatible with a former one mis >>> Warning 1076 The gtid domain being deleted ('1') is not in the current binlog state >>> FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); >>> ERROR HY000: Could not delete gtid domain. Reason: binlog files may contain gtids from the domain ('1') being deleted. Make sure to first purge those files. >>> +MDEV-31140: Missing error from DELETE_DOMAIN_ID when gtid_binlog_state partially matches GTID_LIST. >>> FLUSH BINARY LOGS; >>> PURGE BINARY LOGS TO 'master-bin.000005'; >>> +SET @@SESSION.gtid_domain_id=8; >>> +SET @@SESSION.server_id=10*8 + 1; >>> +INSERT INTO t SELECT 1+MAX(a) FROM t; >>> +FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); >>> +ERROR HY000: Could not delete gtid domain. Reason: binlog files may contain gtids from the domain ('8') being deleted. Make sure to first purge those files. >>> +FLUSH BINARY LOGS; >>> +PURGE BINARY LOGS TO 'master-bin.000006'; >>> FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); >>> Warnings: >>> Warning 1076 The gtid domain being deleted ('0') is not in the current binlog state >>> Gtid_list of the current binlog does not contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0: >>> -show binlog events in 'master-bin.000006' limit 1,1; >>> +show binlog events in 'master-bin.000007' limit 1,1; >>> Log_name Pos Event_type Server_id End_log_pos Info >>> -master-bin.000006 # Gtid_list 1 # [] >>> +master-bin.000007 # Gtid_list 1 # [] >>> SET @@SESSION.gtid_domain_id=1;; >>> SET @@SESSION.server_id=1; >>> SET @@SESSION.gtid_seq_no=1; >>> @@ -75,7 +83,7 @@ INSERT INTO t SET a=1; >>> SELECT @gtid_binlog_state_saved "as original state", @@GLOBAL.gtid_binlog_state as "out of order for 11 domain state"; >>> as original state out of order for 11 domain state >>> 1-1-1,1-2-2,11-11-11 1-1-1,1-2-2,11-11-1 >>> -PURGE BINARY LOGS TO 'master-bin.000007'; >>> +PURGE BINARY LOGS TO 'master-bin.000008'; >>> the following command succeeds with warnings >>> FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1); >>> Warnings: >>> diff --git a/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test b/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test >>> index 8311f4bd800..1643ecff72d 100644 >>> --- a/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test >>> +++ b/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test >>> @@ -21,7 +21,6 @@ FLUSH BINARY LOGS DELETE_DOMAIN_ID = (); >>> --echo but with a warning >>> --let $binlog_pre_flush=query_get_value(SHOW MASTER STATUS, Position, 1) >>> FLUSH BINARY LOGS DELETE_DOMAIN_ID = (99); >>> ---let $binlog_start=$binlog_pre_flush >>> --source include/show_binary_logs.inc >>> >>> # Log one event in a specified domain and try to delete the domain >>> @@ -62,6 +61,8 @@ FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1); >>> # expected overrun of the static buffers of underlying dynamic arrays is doing. >>> --let $domain_cnt=17 >>> --let $server_in_domain_cnt=3 >>> +--let $err_domain_id=`SELECT FLOOR($domain_cnt/2)` >>> +--let $err_server_id=`SELECT FLOOR($server_in_domain_cnt/2)` >>> --let $domain_list= >>> --disable_query_log >>> while ($domain_cnt) >>> @@ -86,6 +87,16 @@ while ($domain_cnt) >>> --error ER_BINLOG_CANT_DELETE_GTID_DOMAIN >>> --eval FLUSH BINARY LOGS DELETE_DOMAIN_ID = ($domain_list) >>> >>> +--echo MDEV-31140: Missing error from DELETE_DOMAIN_ID when gtid_binlog_state partially matches GTID_LIST. >>> +FLUSH BINARY LOGS; >>> +--let $purge_to_binlog= query_get_value(SHOW MASTER STATUS, File, 1) >>> +--eval PURGE BINARY LOGS TO '$purge_to_binlog' >>> +--eval SET @@SESSION.gtid_domain_id=$err_domain_id >>> +--eval SET @@SESSION.server_id=10*$err_domain_id + $err_server_id >>> +eval INSERT INTO t SELECT 1+MAX(a) FROM t; >>> +--error ER_BINLOG_CANT_DELETE_GTID_DOMAIN >>> +--eval FLUSH BINARY LOGS DELETE_DOMAIN_ID = ($domain_list) >>> + >>> # Now satisfy the safety condtion to purge log files containing $domain list >>> FLUSH BINARY LOGS; >>> --let $purge_to_binlog= query_get_value(SHOW MASTER STATUS, File, 1) >>> diff --git a/sql/rpl_gtid.cc b/sql/rpl_gtid.cc >>> index c4e5c75b10a..7b67a83b3dd 100644 >>> --- a/sql/rpl_gtid.cc >>> +++ b/sql/rpl_gtid.cc >>> @@ -2209,18 +2209,16 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, >>> >>> /* >>> For each domain_id from ids >>> - when no such domain in binlog state >>> - warn && continue >>> - For each domain.server's last gtid >>> - when not locate the last gtid in glev.list >>> - error out binlog state can't change >>> - otherwise continue >>> + If the domain is already absent from the binlog state >>> + Warn && continue >>> + If any GTID with that domain in binlog state is missing from glev.list >>> + Error out binlog state can't change >>> */ >>> for (ulong i= 0; i < ids->elements; i++) >>> { >>> rpl_binlog_state::element *elem= NULL; >>> uint32 *ptr_domain_id; >>> - bool not_match; >>> + bool all_found; >>> >>> ptr_domain_id= (uint32*) dynamic_array_ptr(ids, i); >>> elem= (rpl_binlog_state::element *) >>> @@ -2235,14 +2233,18 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, >>> continue; >>> } >>> >>> - for (not_match= true, k= 0; k < elem->hash.records; k++) >>> + all_found= true; >>> + for (k= 0; k < elem->hash.records && all_found; k++) >>> { >>> rpl_gtid *d_gtid= (rpl_gtid *)my_hash_element(&elem->hash, k); >>> - for (ulong l= 0; l < glev->count && not_match; l++) >>> - not_match= !(*d_gtid == glev->list[l]); >>> + bool match_found= false; >>> + for (ulong l= 0; l < glev->count && !match_found; l++) >>> + match_found= match_found || (*d_gtid == glev->list[l]); >>> + if (!match_found) >>> + all_found= false; >>> } >>> >>> - if (not_match) >>> + if (!all_found) >>> { >>> sprintf(errbuf, "binlog files may contain gtids from the domain ('%u') " >>> "being deleted. Make sure to first purge those files", > > modified sql/rpl_gtid.cc > @@ -2220,7 +2220,7 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, > { > rpl_binlog_state::element *elem= NULL; > uint32 *ptr_domain_id; > - bool not_match; > + ulong match; // count-down for matches found > > ptr_domain_id= (uint32*) dynamic_array_ptr(ids, i); > elem= (rpl_binlog_state::element *) > @@ -2235,14 +2235,15 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, > continue; > } > > - for (not_match= true, k= 0; k < elem->hash.records; k++) > + for (match= elem->hash.records, k= 0; > + k < elem->hash.records && match + k == elem->hash.records; k++) > { > rpl_gtid *d_gtid= (rpl_gtid *)my_hash_element(&elem->hash, k); > - for (ulong l= 0; l < glev->count && not_match; l++) > - not_match= !(*d_gtid == glev->list[l]); > + for (ulong l= 0, same= 0; l < glev->count && !same; l++) > + match-= (same= (*d_gtid == glev->list[l])); > } > > - if (not_match) > + if (match) > { > sprintf(errbuf, "binlog files may contain gtids from the domain ('%u') " > "being deleted. Make sure to first purge those files",

1 0

Re: [Maria-developers] MDEV-26301: Obvious wrong query result
by Sergey Petrunia 28 Apr '23

28 Apr '23

Hello Igor, I'm looking at your latest patch for MDEV-26301 in bb-10.4-igor and I see an obvious bug in the execution part: Split-materialized table re-fill is triggered by this code in sub_select(): if (join_tab->split_derived_to_update && !end_of_records) { table_map tab_map= join_tab->split_derived_to_update; for (uint i= 0; tab_map; i++, tab_map>>= 1) { if (tab_map & 1) join->map2table[i]->preread_init_done= false; } } (did you see Table_map_iterator class btw?) but what if the table in question uses join buffer and so uses sub_select_cache(), not sub_select()? Please find the testcase below: # 5 values create table t1(a int, b int); insert into t1 select seq,seq from seq_1_to_5; # 5 value groups of size 2 each create table t2(a int, b int, key(a)); insert into t2 select A.seq,B.seq from seq_1_to_25 A, seq_1_to_2 B; # 5 value groups of size 3 each create table t3(a int, b int, key(a)); insert into t3 select A.seq,B.seq from seq_1_to_5 A, seq_1_to_3 B; analyze table t1,t2,t3 persistent for all; create table t10 ( grp_id int, col1 int, key(grp_id) ); # 100 groups of 100 values each insert into t10 select A.seq, B.seq from seq_1_to_100 A, seq_1_to_100 B; # and X10 multiplier create table t11 ( col1 int, col2 int ); insert into t11 select A.seq, A.seq from seq_1_to_10 A; analyze table t10,t11 persistent for all; explain select * from ( (t1 left join t2 on t2.a=t1.b) left join t3 on t3.a=t1.b ) left join (select grp_id, count(*) from t10 left join t11 on t11.col1=t10.col1 group by grp_id) T on T.grp_id=t1.b; +------+-----------------+------------+------+---------------+--------+---------+---------+------+-------------------------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +------+-----------------+------------+------+---------------+--------+---------+---------+------+-------------------------------------------------+ | 1 | PRIMARY | t1 | ALL | NULL | NULL | NULL | NULL | 5 | | | 1 | PRIMARY | t2 | ref | a | a | 5 | j1.t1.b | 2 | Using where | | 1 | PRIMARY | t3 | ref | a | a | 5 | j1.t1.b | 3 | Using where | | 1 | PRIMARY | <derived2> | ref | key0 | key0 | 5 | j1.t1.b | 10 | Using where | | 2 | LATERAL DERIVED | t10 | ref | grp_id | grp_id | 5 | j1.t1.b | 100 | | | 2 | LATERAL DERIVED | t11 | ALL | NULL | NULL | NULL | NULL | 10 | Using where; Using join buffer (flat, BNL join) | +------+-----------------+------------+------+---------------+--------+---------+---------+------+-------------------------------------------------+ Ok Note the query result: MariaDB [j1]> select * from -> ( -> (t1 left join t2 on t2.a=t1.b) -> left join t3 on t3.a=t1.b -> ) left join (select grp_id, count(*) -> from t10 left join t11 on t11.col1=t10.col1 -> group by grp_id) T on T.grp_id=t1.b; +------+------+------+------+------+------+--------+----------+ | a | b | a | b | a | b | grp_id | count(*) | +------+------+------+------+------+------+--------+----------+ | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 100 | | 1 | 1 | 1 | 1 | 1 | 2 | 1 | 100 | | 1 | 1 | 1 | 1 | 1 | 3 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 3 | 1 | 100 | | 2 | 2 | 2 | 1 | 2 | 1 | 2 | 100 | | 2 | 2 | 2 | 1 | 2 | 2 | 2 | 100 | | 2 | 2 | 2 | 1 | 2 | 3 | 2 | 100 | | 2 | 2 | 2 | 2 | 2 | 1 | 2 | 100 | | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 100 | | 2 | 2 | 2 | 2 | 2 | 3 | 2 | 100 | | 3 | 3 | 3 | 1 | 3 | 1 | 3 | 100 | | 3 | 3 | 3 | 1 | 3 | 2 | 3 | 100 | | 3 | 3 | 3 | 1 | 3 | 3 | 3 | 100 | | 3 | 3 | 3 | 2 | 3 | 1 | 3 | 100 | | 3 | 3 | 3 | 2 | 3 | 2 | 3 | 100 | | 3 | 3 | 3 | 2 | 3 | 3 | 3 | 100 | | 4 | 4 | 4 | 1 | 4 | 1 | 4 | 100 | | 4 | 4 | 4 | 1 | 4 | 2 | 4 | 100 | | 4 | 4 | 4 | 1 | 4 | 3 | 4 | 100 | | 4 | 4 | 4 | 2 | 4 | 1 | 4 | 100 | | 4 | 4 | 4 | 2 | 4 | 2 | 4 | 100 | | 4 | 4 | 4 | 2 | 4 | 3 | 4 | 100 | | 5 | 5 | 5 | 1 | 5 | 1 | 5 | 100 | | 5 | 5 | 5 | 1 | 5 | 2 | 5 | 100 | | 5 | 5 | 5 | 1 | 5 | 3 | 5 | 100 | | 5 | 5 | 5 | 2 | 5 | 1 | 5 | 100 | | 5 | 5 | 5 | 2 | 5 | 2 | 5 | 100 | | 5 | 5 | 5 | 2 | 5 | 3 | 5 | 100 | +------+------+------+------+------+------+--------+----------+ 30 rows in set (0.015 sec) Now, let's make the tables not use indexes and use join buffer in the top-level select: explain select * from ( (t1 left join t2 use index() on t2.a=t1.b) left join t3 use index() on t3.a=t1.b ) left join (select grp_id, count(*) from t10 left join t11 on t11.col1=t10.col1 group by grp_id) T on T.grp_id=t1.b; +------+-----------------+------------+------+---------------+--------+---------+---------+------+--------------------------------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +------+-----------------+------------+------+---------------+--------+---------+---------+------+--------------------------------------------------------+ | 1 | PRIMARY | t1 | ALL | NULL | NULL | NULL | NULL | 5 | | | 1 | PRIMARY | t2 | ALL | NULL | NULL | NULL | NULL | 50 | Using where; Using join buffer (flat, BNL join) | | 1 | PRIMARY | t3 | ALL | NULL | NULL | NULL | NULL | 15 | Using where; Using join buffer (incremental, BNL join) | | 1 | PRIMARY | <derived2> | ref | key0 | key0 | 5 | j1.t1.b | 10 | Using where | | 2 | LATERAL DERIVED | t10 | ref | grp_id | grp_id | 5 | j1.t1.b | 100 | | | 2 | LATERAL DERIVED | t11 | ALL | NULL | NULL | NULL | NULL | 10 | Using where; Using join buffer (flat, BNL join) | +------+-----------------+------------+------+---------------+--------+---------+---------+------+--------------------------------------------------------+ select * from ( (t1 left join t2 use index() on t2.a=t1.b) left join t3 use index() on t3.a=t1.b ) left join (select grp_id, count(*) from t10 left join t11 on t11.col1=t10.col1 group by grp_id) T on T.grp_id=t1.b; +------+------+------+------+------+------+--------+----------+ | a | b | a | b | a | b | grp_id | count(*) | +------+------+------+------+------+------+--------+----------+ | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 100 | | 1 | 1 | 1 | 1 | 1 | 2 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 100 | | 1 | 1 | 1 | 1 | 1 | 3 | 1 | 100 | | 1 | 1 | 1 | 2 | 1 | 3 | 1 | 100 | | 2 | 2 | 2 | 1 | 2 | 1 | NULL | NULL | | 2 | 2 | 2 | 2 | 2 | 1 | NULL | NULL | | 2 | 2 | 2 | 1 | 2 | 2 | NULL | NULL | | 2 | 2 | 2 | 2 | 2 | 2 | NULL | NULL | | 2 | 2 | 2 | 1 | 2 | 3 | NULL | NULL | | 2 | 2 | 2 | 2 | 2 | 3 | NULL | NULL | | 3 | 3 | 3 | 1 | 3 | 1 | NULL | NULL | | 3 | 3 | 3 | 2 | 3 | 1 | NULL | NULL | | 3 | 3 | 3 | 1 | 3 | 2 | NULL | NULL | | 3 | 3 | 3 | 2 | 3 | 2 | NULL | NULL | | 3 | 3 | 3 | 1 | 3 | 3 | NULL | NULL | | 3 | 3 | 3 | 2 | 3 | 3 | NULL | NULL | | 4 | 4 | 4 | 1 | 4 | 1 | NULL | NULL | | 4 | 4 | 4 | 2 | 4 | 1 | NULL | NULL | | 4 | 4 | 4 | 1 | 4 | 2 | NULL | NULL | | 4 | 4 | 4 | 2 | 4 | 2 | NULL | NULL | | 4 | 4 | 4 | 1 | 4 | 3 | NULL | NULL | | 4 | 4 | 4 | 2 | 4 | 3 | NULL | NULL | | 5 | 5 | 5 | 1 | 5 | 1 | NULL | NULL | | 5 | 5 | 5 | 2 | 5 | 1 | NULL | NULL | | 5 | 5 | 5 | 1 | 5 | 2 | NULL | NULL | | 5 | 5 | 5 | 2 | 5 | 2 | NULL | NULL | | 5 | 5 | 5 | 1 | 5 | 3 | NULL | NULL | | 5 | 5 | 5 | 2 | 5 | 3 | NULL | NULL | +------+------+------+------+------+------+--------+----------+ 30 rows in set (0.008 sec) Bummer. BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net

1 0

Re: [Maria-developers] d25439f1cbb: MDEV-31140: FLUSH BINARY LOGS DELETE_DOMAIN_ID=(D) can errorneously delete active domains
by Kristian Nielsen 27 Apr '23

27 Apr '23

Hi Andrei, Want to review a small patch for https://jira.mariadb.org/browse/MDEV-31140 ? I found this problem while trying to find the root cause of https://jira.mariadb.org/browse/MDEV-30386 . However this bug seems to be unrelated to MDEV-30386. Basically the logic for comparing the gtid_binlog_state with the GTID_LIST event was flawed, making it possible to delete a domain_id from the binlog state when this should not be allowed. - Kristian. Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: > revision-id: d25439f1cbb79e5467b4249792130bfe524e3f10 (mariadb-10.11.2-21-gd25439f1cbb) > parent(s): 656c2e18b1e9ea5d0314745f3988d126eedbc22a > author: Kristian Nielsen > committer: Kristian Nielsen > timestamp: 2023-04-27 12:16:07 +0200 > message: > > MDEV-31140: FLUSH BINARY LOGS DELETE_DOMAIN_ID=(D) can errorneously delete active domains > > Fix the code in rpl_binlog_state::drop_domain(), so that _all_ entries for > the domain in the binlog state must match an entry in the initial GTID_LIST, > not just one entry match. > > Signed-off-by: Kristian Nielsen <knielsen(a)knielsen-hq.org> > > --- > .../r/binlog_flush_binlogs_delete_domain.result | 14 ++++++++++--- > .../t/binlog_flush_binlogs_delete_domain.test | 13 +++++++++++- > sql/rpl_gtid.cc | 24 ++++++++++++---------- > 3 files changed, 36 insertions(+), 15 deletions(-) > > diff --git a/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result b/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result > index fdcfb4bfa01..1c11191802f 100644 > --- a/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result > +++ b/mysql-test/suite/binlog/r/binlog_flush_binlogs_delete_domain.result > @@ -46,15 +46,23 @@ Warning 1076 The current gtid binlog state is incompatible with a former one mis > Warning 1076 The gtid domain being deleted ('1') is not in the current binlog state > FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); > ERROR HY000: Could not delete gtid domain. Reason: binlog files may contain gtids from the domain ('1') being deleted. Make sure to first purge those files. > +MDEV-31140: Missing error from DELETE_DOMAIN_ID when gtid_binlog_state partially matches GTID_LIST. > FLUSH BINARY LOGS; > PURGE BINARY LOGS TO 'master-bin.000005'; > +SET @@SESSION.gtid_domain_id=8; > +SET @@SESSION.server_id=10*8 + 1; > +INSERT INTO t SELECT 1+MAX(a) FROM t; > +FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); > +ERROR HY000: Could not delete gtid domain. Reason: binlog files may contain gtids from the domain ('8') being deleted. Make sure to first purge those files. > +FLUSH BINARY LOGS; > +PURGE BINARY LOGS TO 'master-bin.000006'; > FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0); > Warnings: > Warning 1076 The gtid domain being deleted ('0') is not in the current binlog state > Gtid_list of the current binlog does not contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 0: > -show binlog events in 'master-bin.000006' limit 1,1; > +show binlog events in 'master-bin.000007' limit 1,1; > Log_name Pos Event_type Server_id End_log_pos Info > -master-bin.000006 # Gtid_list 1 # [] > +master-bin.000007 # Gtid_list 1 # [] > SET @@SESSION.gtid_domain_id=1;; > SET @@SESSION.server_id=1; > SET @@SESSION.gtid_seq_no=1; > @@ -75,7 +83,7 @@ INSERT INTO t SET a=1; > SELECT @gtid_binlog_state_saved "as original state", @@GLOBAL.gtid_binlog_state as "out of order for 11 domain state"; > as original state out of order for 11 domain state > 1-1-1,1-2-2,11-11-11 1-1-1,1-2-2,11-11-1 > -PURGE BINARY LOGS TO 'master-bin.000007'; > +PURGE BINARY LOGS TO 'master-bin.000008'; > the following command succeeds with warnings > FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1); > Warnings: > diff --git a/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test b/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test > index 8311f4bd800..1643ecff72d 100644 > --- a/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test > +++ b/mysql-test/suite/binlog/t/binlog_flush_binlogs_delete_domain.test > @@ -21,7 +21,6 @@ FLUSH BINARY LOGS DELETE_DOMAIN_ID = (); > --echo but with a warning > --let $binlog_pre_flush=query_get_value(SHOW MASTER STATUS, Position, 1) > FLUSH BINARY LOGS DELETE_DOMAIN_ID = (99); > ---let $binlog_start=$binlog_pre_flush > --source include/show_binary_logs.inc > > # Log one event in a specified domain and try to delete the domain > @@ -62,6 +61,8 @@ FLUSH BINARY LOGS DELETE_DOMAIN_ID = (1); > # expected overrun of the static buffers of underlying dynamic arrays is doing. > --let $domain_cnt=17 > --let $server_in_domain_cnt=3 > +--let $err_domain_id=`SELECT FLOOR($domain_cnt/2)` > +--let $err_server_id=`SELECT FLOOR($server_in_domain_cnt/2)` > --let $domain_list= > --disable_query_log > while ($domain_cnt) > @@ -86,6 +87,16 @@ while ($domain_cnt) > --error ER_BINLOG_CANT_DELETE_GTID_DOMAIN > --eval FLUSH BINARY LOGS DELETE_DOMAIN_ID = ($domain_list) > > +--echo MDEV-31140: Missing error from DELETE_DOMAIN_ID when gtid_binlog_state partially matches GTID_LIST. > +FLUSH BINARY LOGS; > +--let $purge_to_binlog= query_get_value(SHOW MASTER STATUS, File, 1) > +--eval PURGE BINARY LOGS TO '$purge_to_binlog' > +--eval SET @@SESSION.gtid_domain_id=$err_domain_id > +--eval SET @@SESSION.server_id=10*$err_domain_id + $err_server_id > +eval INSERT INTO t SELECT 1+MAX(a) FROM t; > +--error ER_BINLOG_CANT_DELETE_GTID_DOMAIN > +--eval FLUSH BINARY LOGS DELETE_DOMAIN_ID = ($domain_list) > + > # Now satisfy the safety condtion to purge log files containing $domain list > FLUSH BINARY LOGS; > --let $purge_to_binlog= query_get_value(SHOW MASTER STATUS, File, 1) > diff --git a/sql/rpl_gtid.cc b/sql/rpl_gtid.cc > index c4e5c75b10a..7b67a83b3dd 100644 > --- a/sql/rpl_gtid.cc > +++ b/sql/rpl_gtid.cc > @@ -2209,18 +2209,16 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, > > /* > For each domain_id from ids > - when no such domain in binlog state > - warn && continue > - For each domain.server's last gtid > - when not locate the last gtid in glev.list > - error out binlog state can't change > - otherwise continue > + If the domain is already absent from the binlog state > + Warn && continue > + If any GTID with that domain in binlog state is missing from glev.list > + Error out binlog state can't change > */ > for (ulong i= 0; i < ids->elements; i++) > { > rpl_binlog_state::element *elem= NULL; > uint32 *ptr_domain_id; > - bool not_match; > + bool all_found; > > ptr_domain_id= (uint32*) dynamic_array_ptr(ids, i); > elem= (rpl_binlog_state::element *) > @@ -2235,14 +2233,18 @@ rpl_binlog_state::drop_domain(DYNAMIC_ARRAY *ids, > continue; > } > > - for (not_match= true, k= 0; k < elem->hash.records; k++) > + all_found= true; > + for (k= 0; k < elem->hash.records && all_found; k++) > { > rpl_gtid *d_gtid= (rpl_gtid *)my_hash_element(&elem->hash, k); > - for (ulong l= 0; l < glev->count && not_match; l++) > - not_match= !(*d_gtid == glev->list[l]); > + bool match_found= false; > + for (ulong l= 0; l < glev->count && !match_found; l++) > + match_found= match_found || (*d_gtid == glev->list[l]); > + if (!match_found) > + all_found= false; > } > > - if (not_match) > + if (!all_found) > { > sprintf(errbuf, "binlog files may contain gtids from the domain ('%u') " > "being deleted. Make sure to first purge those files",

1 0

Re: [Maria-developers] MDEV-30946 Index usage for DATE(datetime_column) = const does not work for DELETE and UPDATE
by Sergey Petrunia 12 Apr '23

12 Apr '23

Hi Oleg, Why is the call to Item::date_conds_transformer done after prune_partitions()? This causes partition pruning not to be able to use the conditions: --source include/have_partition.inc --source include/have_sequence.inc create table t1 ( a datetime, key(a) ) partition by range(year(a)) ( partition p0 values less than (2022), partition p1 values less than (MAXVALUE) ); insert into t1 select date_add('2020-01-01', interval seq day) from seq_1_to_1000; explain partitions select * from t1 where year(a) = 2020; explain partitions delete from t1 where year(a) = 2020; drop table t1; Gives: explain partitions select * from t1 where year(a) = 2020; id select_type table partitions type possible_keys key key_len ref rows Extra 1 SIMPLE t1 p0 range a a 6 NULL 348 Using where; Using index explain partitions delete from t1 where year(a) = 2020; id select_type table partitions type possible_keys key key_len ref rows Extra 1 SIMPLE t1 p0,p1 ALL a NULL NULL NULL 1000 Using where Note that SELECT uses one partition while DELETE uses two. Please fix this. BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net

1 0

[Maria-developers] Participation in MariaDB Community's Open Source Project
by 和映泉 09 Apr '23

09 Apr '23

Dear MariaDB Development Team, I am writing to express my interest in participating in your open source project. I am an experienced programmer with knowledge of multiple programming languages and technologies, and I have some understanding of MariaDB products and source code. I believe I can make positive contributions to your project and gain more learning and growth opportunities from it. Here is a summary of my programming experience and skills: - I have 6 years of experience in software development and have participated in commercial projects. - I am proficient in multiple programming languages such as Java, Python, and Scala, and have knowledge of SQL and database technologies. - I am familiar with Linux systems and command-line tools, and can use version control tools such as Git. - I have some understanding and practical experience in software testing, code quality, and performance optimization. In addition, I have also used MySQL products and have some understanding of their functions and features. I believe this will help me better understand and contribute to your open source project. Thank you for taking the time to read my email. I look forward to your reply and hope to have the opportunity to contribute to the MariaDB community. Best regards, Yingquan He

2 1

[Maria-developers] CI failing on almost all branches - time to adopt protected branches?
by Otto Kekäläinen 03 Apr '23

03 Apr '23

Hello! Has there recently been any discussions about enabling the Github 'protected branch' feature? Review of current mainlines shows that ALL but 10.4 are broken and most of them on different tests. I am afraid that quality and test fixes will never catch up if people are allowed to push code that makes testable things fail. Using protected branches would prevent people from being able to push commits on mainline that fail in CI. https://github.com/MariaDB/server/tree/10.3 / 4c4939b NOT OK: - amd64-debian-10-debug-embedded fails on sql_sequence.kill - amd64-fedora-36 fails on main.func_math https://github.com/MariaDB/server/tree/10.4 / 1767390 OK https://github.com/MariaDB/server/tree/10.5 / ac5a534 NOT OK: - amd64-ubuntu-2004-msan fails on main.selectivity_innodb - amd64-windows fails to compile https://github.com/MariaDB/server/tree/10.6 / dc1d621 NOT OK: - buildbot/amd64-ubuntu-2204-msan fails on innodb.gap_locks https://github.com/MariaDB/server/tree/10.7 / bc656c4 NOT OK: - buildbot/amd64-debian-10 fails to start MTR https://github.com/MariaDB/server/tree/10.8 / dd2fe81 NOT OK: - amd64-ubuntu-2204-msan fails on MTR timeout (crash?) https://github.com/MariaDB/server/tree/10.9 / 55e78eb NOT OK: - buildbot/amd64-ubuntu-2204-msan fails on mariabackup.incremental_page_compressed innodb.innodb_page_compressed https://github.com/MariaDB/server/tree/10.10 / 191821f NOT OK: - amd64-ubuntu-2004-debug fails on binlog.binlog_truncate_multi_engine - amd64-ubuntu-2204-debug-ps fails on main.bad_startup_options https://github.com/MariaDB/server/tree/10.11 / d84a282 NOT OK: - buildbot/amd64-ubuntu-2204-msan fails on mariabackup.incremental_page_compressed innodb.innodb_page_compressed https://github.com/MariaDB/server/tree/11.0 / 8e55d7e NOT OK: - buildbot/amd64-ubuntu-2004-msan fails on innodb.table_definition_cache_debug main.lock_sync - continuous-integration/appveyor/branch fails on main.innodb_ext_key

2 1

[Maria-developers] How to get started with GSOC contribution
by Komala Elusuri 25 Mar '23

25 Mar '23

Respected sir, I am Komala, a computer science undergraduate from India. I am new to open source contributions but I am well aware of python, C, C++, Java,JavaScript and Databases. I did certifications on Google cloud computing foundation with Kubernetes and Python projects from Infosys springboard. I have experience as an intern from Smart Knower. I would love to contribute to your organisation but could you please tell me how to get started? Hoping to hear from you soon. Regards Komala

1 0

Re: [Maria-developers] 940d028521f: MDEV-30164 System variable for default collations
by Sergei Golubchik 21 Mar '23

21 Mar '23

Hi, Alexander, I'm sorry it took a while. I'm still thinking about the whole thing, it's a rather big change for a really fringe functionality. But I failed to come up with something better so far. Code-wise the patch is mostly fine. See few small comments below, and one slightly larger comment re. replication. On Mar 07, Alexander Barkov wrote: > revision-id: 940d028521f (mariadb-10.11.1-4-g940d028521f) > parent(s): 71dedd0ebcd > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2022-12-14 20:00:22 +0400 > message: > > MDEV-30164 System variable for default collations > > This patch adds a way to override default collations > (or "character set collations") for desired character sets. > > diff --git a/sql/lex_charset.h b/sql/lex_charset.h > --- a/sql/lex_charset.h > +++ b/sql/lex_charset.h > @@ -544,6 +699,20 @@ struct Lex_exact_charset_extended_collation_attrs_st > { > return m_ci; > } > + CHARSET_INFO *charset_info(const Charset_collation_map_st &map) const > + { > + switch (m_type) > + { > + case Lex_exact_charset_extended_collation_attrs_st::TYPE_CHARACTER_SET: > + return map.get_collation_for_charset(m_ci); > + case TYPE_EMPTY: Lex_exact_charset_extended_collation_attrs_st::TYPE_EMPTY (or all case labels without a struct name) > + case Lex_exact_charset_extended_collation_attrs_st::TYPE_CHARACTER_SET_COLLATE_EXACT: > + case Lex_exact_charset_extended_collation_attrs_st::TYPE_COLLATE_CONTEXTUALLY_TYPED: COLLATE DEFAULT is TYPE_COLLATE_CONTEXTUALLY_TYPED. shouldn't it use the map too? > + case Lex_exact_charset_extended_collation_attrs_st::TYPE_COLLATE_EXACT: > + break; > + } > + return m_ci; > + } > Type type() const > { > return m_type; > diff --git a/sql/log_event.h b/sql/log_event.h > --- a/sql/log_event.h > +++ b/sql/log_event.h > @@ -2147,6 +2153,8 @@ class Query_log_event: public Log_event > bool sql_mode_inited; > bool charset_inited; > > + LEX_CSTRING character_set_collations_str; better LEX_CUSTRING, don't you think? It's not even a _str. > + > uint32 flags2; > sql_mode_t sql_mode; > ulong auto_increment_increment, auto_increment_offset; > diff --git a/sql/log_event_server.cc b/sql/log_event_server.cc > --- a/sql/log_event_server.cc > +++ b/sql/log_event_server.cc > @@ -1194,6 +1194,14 @@ bool Query_log_event::write() > int2store(start+2, auto_increment_offset); > start+= 4; > } > + > + if (thd && thd->variables.character_set_collations.count()) > + { > + *start++= Q_COLLATIONS_SESSION; > + size_t len= thd->variables.character_set_collations.to_binary((char*)start); > + start+= len; > + } Perhaps, detect if it's needed? A cheap way of doing it would be to extend your Elem_st with a query_id. And every time you find_elem_by_charset, you set this elem's query_id to thd->query_id. And here you write only elements with the current query id. If any. Another approach would be to have a bitmap, like uchar used_default_coll_mapping; and in find_elem_by_charset() you set the bit, like used_default_coll_mapping |= 1 << i; and then, again, print affected collations, if any. Most often used_default_coll_mapping will likely be zero one more question. In, say, 10.10->11.1 replication master and slave will have different default collations, but thd->variables.character_set_collations will not reflect that. How do you plan to solve it? > + > if (charset_inited) > { > *start++= Q_CHARSET_CODE; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 3

Re: [Maria-developers] MDEV-28152: Features for sequence
by Oleksandr Byelkin 21 Mar '23

21 Mar '23

Hi, Yuchen! As for first work for the server itself I find it is very good, but please read my notes below (many of them just code style ones) JFYI if you fix something in the wrong or old code style, you should fix it, but please do not fix the style only things in the code you are not going to change. It is one of the not yet written ( :( ) rules of our fix code guide. As well as code style defined only by old memory and examples in the code. > diff --git a/mysql-test/main/information_schema.result b/mysql-test/main/information_schema.result > index e46014e44b9..ce111c58796 100644 > --- a/mysql-test/main/information_schema.result > +++ b/mysql-test/main/information_schema.result > @@ -81,6 +81,7 @@ REFERENTIAL_CONSTRAINTS > ROUTINES > SCHEMATA > SCHEMA_PRIVILEGES > +SEQUENCES > SESSION_STATUS > SESSION_VARIABLES > SPATIAL_REF_SYS > diff --git a/mysql-test/main/information_schema_all_engines.result b/mysql-test/main/information_schema_all_engines.result > index 23a853e363c..316871995c1 100644 > --- a/mysql-test/main/information_schema_all_engines.result > +++ b/mysql-test/main/information_schema_all_engines.result > @@ -52,6 +52,7 @@ REFERENTIAL_CONSTRAINTS > ROUTINES > SCHEMATA > SCHEMA_PRIVILEGES > +SEQUENCES > SESSION_STATUS > SESSION_VARIABLES > SPATIAL_REF_SYS > @@ -133,6 +134,7 @@ REFERENTIAL_CONSTRAINTS CONSTRAINT_SCHEMA > ROUTINES ROUTINE_SCHEMA > SCHEMATA SCHEMA_NAME > SCHEMA_PRIVILEGES TABLE_SCHEMA > +SEQUENCES SEQUENCE_SCHEMA > SESSION_STATUS VARIABLE_NAME > SESSION_VARIABLES VARIABLE_NAME > SPATIAL_REF_SYS SRID > @@ -214,6 +216,7 @@ REFERENTIAL_CONSTRAINTS CONSTRAINT_SCHEMA > ROUTINES ROUTINE_SCHEMA > SCHEMATA SCHEMA_NAME > SCHEMA_PRIVILEGES TABLE_SCHEMA > +SEQUENCES SEQUENCE_SCHEMA > SESSION_STATUS VARIABLE_NAME > SESSION_VARIABLES VARIABLE_NAME > SPATIAL_REF_SYS SRID > @@ -299,6 +302,7 @@ REFERENTIAL_CONSTRAINTS information_schema.REFERENTIAL_CONSTRAINTS 1 > ROUTINES information_schema.ROUTINES 1 > SCHEMATA information_schema.SCHEMATA 1 > SCHEMA_PRIVILEGES information_schema.SCHEMA_PRIVILEGES 1 > +SEQUENCES information_schema.SEQUENCES 1 > SESSION_STATUS information_schema.SESSION_STATUS 1 > SESSION_VARIABLES information_schema.SESSION_VARIABLES 1 > SPATIAL_REF_SYS information_schema.SPATIAL_REF_SYS 1 > @@ -369,6 +373,7 @@ Database: information_schema > | ROUTINES | > | SCHEMATA | > | SCHEMA_PRIVILEGES | > +| SEQUENCES | > | SESSION_STATUS | > | SESSION_VARIABLES | > | SPATIAL_REF_SYS | > @@ -440,6 +445,7 @@ Database: INFORMATION_SCHEMA > | ROUTINES | > | SCHEMATA | > | SCHEMA_PRIVILEGES | > +| SEQUENCES | > | SESSION_STATUS | > | SESSION_VARIABLES | > | SPATIAL_REF_SYS | > @@ -463,5 +469,5 @@ Wildcard: inf_rmation_schema > | information_schema | > SELECT table_schema, count(*) FROM information_schema.TABLES WHERE table_schema IN ('mysql', 'INFORMATION_SCHEMA', 'test', 'mysqltest') GROUP BY TABLE_SCHEMA; > table_schema count(*) > -information_schema 66 > +information_schema 67 > mysql 31 > diff --git a/mysql-test/main/mysqldump.result b/mysql-test/main/mysqldump.result > index 44ed94e43c9..41607ea3068 100644 > --- a/mysql-test/main/mysqldump.result > +++ b/mysql-test/main/mysqldump.result > @@ -6369,16 +6369,16 @@ NEXTVAL(d.s1) NEXTVAL(d.s2) NEXTVAL(d.s3) NEXTVAL(d.s4) > # Show create before dump > show create sequence d.s1; > Table Create Table > -s1 CREATE SEQUENCE `s1` start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > +s1 CREATE SEQUENCE `s1` as bigint start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s2; > Table Create Table > -s2 CREATE SEQUENCE `s2` start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > +s2 CREATE SEQUENCE `s2` as bigint start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s3; > Table Create Table > -s3 CREATE SEQUENCE `s3` start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > +s3 CREATE SEQUENCE `s3` as bigint start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s4; > Table Create Table > -s4 CREATE SEQUENCE `s4` start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > +s4 CREATE SEQUENCE `s4` as bigint start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > # Dump sequence without `--no-data` > # Restore from mysqldump > SETVAL(`s1`, 1101, 0) > @@ -6392,16 +6392,16 @@ SETVAL(`s4`, 1401, 0) > # Show create after restore > show create sequence d.s1; > Table Create Table > -s1 CREATE SEQUENCE `s1` start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > +s1 CREATE SEQUENCE `s1` as bigint start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s2; > Table Create Table > -s2 CREATE SEQUENCE `s2` start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > +s2 CREATE SEQUENCE `s2` as bigint start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s3; > Table Create Table > -s3 CREATE SEQUENCE `s3` start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > +s3 CREATE SEQUENCE `s3` as bigint start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s4; > Table Create Table > -s4 CREATE SEQUENCE `s4` start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > +s4 CREATE SEQUENCE `s4` as bigint start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > SELECT NEXTVAL(d.s1),NEXTVAL(d.s2),NEXTVAL(d.s3), NEXTVAL(d.s4); > NEXTVAL(d.s1) NEXTVAL(d.s2) NEXTVAL(d.s3) NEXTVAL(d.s4) > 100 200 300 400 > @@ -6418,16 +6418,16 @@ SETVAL(`s4`, 1401, 0) > # Show create after restore `--no-data` > show create sequence d.s1; > Table Create Table > -s1 CREATE SEQUENCE `s1` start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > +s1 CREATE SEQUENCE `s1` as bigint start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s2; > Table Create Table > -s2 CREATE SEQUENCE `s2` start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > +s2 CREATE SEQUENCE `s2` as bigint start with 200 minvalue 200 maxvalue 1200 increment by 20 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s3; > Table Create Table > -s3 CREATE SEQUENCE `s3` start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > +s3 CREATE SEQUENCE `s3` as bigint start with 300 minvalue 300 maxvalue 1300 increment by 30 cache 1000 cycle ENGINE=MyISAM > show create sequence d.s4; > Table Create Table > -s4 CREATE SEQUENCE `s4` start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > +s4 CREATE SEQUENCE `s4` as bigint start with 400 minvalue 400 maxvalue 1400 increment by 40 cache 1000 cycle ENGINE=MyISAM > SELECT NEXTVAL(d.s1),NEXTVAL(d.s2),NEXTVAL(d.s3), NEXTVAL(d.s4); > NEXTVAL(d.s1) NEXTVAL(d.s2) NEXTVAL(d.s3) NEXTVAL(d.s4) > 100 200 300 400 > @@ -6443,7 +6443,7 @@ SETVAL(`s4`, 1401, 0) > 1401 > show create sequence d2.s1; > Table Create Table > -s1 CREATE SEQUENCE `s1` start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > +s1 CREATE SEQUENCE `s1` as bigint start with 100 minvalue 100 maxvalue 1100 increment by 10 cache 1000 cycle ENGINE=MyISAM > drop sequence d.s1, d.s2, d.s3, d.s4; > drop database d; > drop database d2; > diff --git a/mysql-test/suite/funcs_1/r/is_columns_is.result b/mysql-test/suite/funcs_1/r/is_columns_is.result > index c88a3a9ac8d..03eece0b4ae 100644 > --- a/mysql-test/suite/funcs_1/r/is_columns_is.result > +++ b/mysql-test/suite/funcs_1/r/is_columns_is.result > @@ -339,6 +339,21 @@ def information_schema SCHEMA_PRIVILEGES IS_GRANTABLE 5 NULL NO varchar 3 9 NULL > def information_schema SCHEMA_PRIVILEGES PRIVILEGE_TYPE 4 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > def information_schema SCHEMA_PRIVILEGES TABLE_CATALOG 2 NULL NO varchar 512 1536 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(512) select NEVER NULL > def information_schema SCHEMA_PRIVILEGES TABLE_SCHEMA 3 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > +def information_schema SEQUENCES CYCLE_OPTION 12 NULL NO bigint NULL NULL 19 0 NULL NULL NULL bigint(21) select NEVER NULL As I remember CYCLE_OPTION is boolean or YES/NO, why does it become bigint ? > +def information_schema SEQUENCES DATA_TYPE 4 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > +def information_schema SEQUENCES DECLARED_DATA_TYPE 13 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL > +def information_schema SEQUENCES DECLARED_NUMERIC_PRECISION 14 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL > +def information_schema SEQUENCES DECLARED_NUMERIC_SCALE 15 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL We do not need DECLARED_* it is only needed for feature T322 “Declared data type attributes” and we do not have it. > +def information_schema SEQUENCES INCREMENT 11 NULL NO bigint NULL NULL 19 0 NULL NULL NULL bigint(21) select NEVER NULL > +def information_schema SEQUENCES MAXIMUM_VALUE 10 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) select NEVER NULL > +def information_schema SEQUENCES MINIMUM_VALUE 9 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) select NEVER NULL good decision to make it decimal > +def information_schema SEQUENCES NUMERIC_PRECISION 5 NULL NO int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL > +def information_schema SEQUENCES NUMERIC_PRECISION_RADIX 6 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL I have not found tests of NUMERIC_* output for all supported by us types. > +def information_schema SEQUENCES NUMERIC_SCALE 7 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) select NEVER NULL > +def information_schema SEQUENCES SEQUENCE_CATALOG 1 NULL NO varchar 512 1536 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(512) select NEVER NULL > +def information_schema SEQUENCES SEQUENCE_NAME 3 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > +def information_schema SEQUENCES SEQUENCE_SCHEMA 2 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > +def information_schema SEQUENCES START_VALUE 8 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) select NEVER NULL I do not see CACHED reflected here, please add it. > def information_schema SESSION_STATUS VARIABLE_NAME 1 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > def information_schema SESSION_STATUS VARIABLE_VALUE 2 NULL NO varchar 2048 6144 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(2048) select NEVER NULL > def information_schema SESSION_VARIABLES VARIABLE_NAME 1 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) select NEVER NULL > @@ -890,6 +905,21 @@ NULL information_schema ROUTINES LAST_ALTERED datetime NULL NULL NULL NULL datet > 3.0000 information_schema SCHEMA_PRIVILEGES TABLE_SCHEMA varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SCHEMA_PRIVILEGES PRIVILEGE_TYPE varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SCHEMA_PRIVILEGES IS_GRANTABLE varchar 3 9 utf8mb3 utf8mb3_general_ci varchar(3) > +3.0000 information_schema SEQUENCES SEQUENCE_CATALOG varchar 512 1536 utf8mb3 utf8mb3_general_ci varchar(512) > +3.0000 information_schema SEQUENCES SEQUENCE_SCHEMA varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +3.0000 information_schema SEQUENCES SEQUENCE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +3.0000 information_schema SEQUENCES DATA_TYPE varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +NULL information_schema SEQUENCES NUMERIC_PRECISION int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES NUMERIC_PRECISION_RADIX int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES NUMERIC_SCALE int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES START_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES MINIMUM_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES MAXIMUM_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES INCREMENT bigint NULL NULL NULL NULL bigint(21) > +NULL information_schema SEQUENCES CYCLE_OPTION bigint NULL NULL NULL NULL bigint(21) > +NULL information_schema SEQUENCES DECLARED_DATA_TYPE int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES DECLARED_NUMERIC_PRECISION int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES DECLARED_NUMERIC_SCALE int NULL NULL NULL NULL int(21) > 3.0000 information_schema SESSION_STATUS VARIABLE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SESSION_STATUS VARIABLE_VALUE varchar 2048 6144 utf8mb3 utf8mb3_general_ci varchar(2048) > 3.0000 information_schema SESSION_VARIABLES VARIABLE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 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 bb12a0c38df..87cc440cdbc 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 > @@ -339,6 +339,21 @@ def information_schema SCHEMA_PRIVILEGES IS_GRANTABLE 5 NULL NO varchar 3 9 NULL > def information_schema SCHEMA_PRIVILEGES PRIVILEGE_TYPE 4 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > def information_schema SCHEMA_PRIVILEGES TABLE_CATALOG 2 NULL NO varchar 512 1536 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(512) NEVER NULL > def information_schema SCHEMA_PRIVILEGES TABLE_SCHEMA 3 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > +def information_schema SEQUENCES CYCLE_OPTION 12 NULL NO bigint NULL NULL 19 0 NULL NULL NULL bigint(21) NEVER NULL > +def information_schema SEQUENCES DATA_TYPE 4 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > +def information_schema SEQUENCES DECLARED_DATA_TYPE 13 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES DECLARED_NUMERIC_PRECISION 14 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES DECLARED_NUMERIC_SCALE 15 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES INCREMENT 11 NULL NO bigint NULL NULL 19 0 NULL NULL NULL bigint(21) NEVER NULL > +def information_schema SEQUENCES MAXIMUM_VALUE 10 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) NEVER NULL > +def information_schema SEQUENCES MINIMUM_VALUE 9 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) NEVER NULL > +def information_schema SEQUENCES NUMERIC_PRECISION 5 NULL NO int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES NUMERIC_PRECISION_RADIX 6 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES NUMERIC_SCALE 7 NULL YES int NULL NULL 10 0 NULL NULL NULL int(21) NEVER NULL > +def information_schema SEQUENCES SEQUENCE_CATALOG 1 NULL NO varchar 512 1536 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(512) NEVER NULL > +def information_schema SEQUENCES SEQUENCE_NAME 3 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > +def information_schema SEQUENCES SEQUENCE_SCHEMA 2 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > +def information_schema SEQUENCES START_VALUE 8 NULL NO decimal NULL NULL 21 0 NULL NULL NULL decimal(21,0) NEVER NULL > def information_schema SESSION_STATUS VARIABLE_NAME 1 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > def information_schema SESSION_STATUS VARIABLE_VALUE 2 NULL NO varchar 2048 6144 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(2048) NEVER NULL > def information_schema SESSION_VARIABLES VARIABLE_NAME 1 NULL NO varchar 64 192 NULL NULL NULL utf8mb3 utf8mb3_general_ci varchar(64) NEVER NULL > @@ -890,6 +905,21 @@ NULL information_schema ROUTINES LAST_ALTERED datetime NULL NULL NULL NULL datet > 3.0000 information_schema SCHEMA_PRIVILEGES TABLE_SCHEMA varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SCHEMA_PRIVILEGES PRIVILEGE_TYPE varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SCHEMA_PRIVILEGES IS_GRANTABLE varchar 3 9 utf8mb3 utf8mb3_general_ci varchar(3) > +3.0000 information_schema SEQUENCES SEQUENCE_CATALOG varchar 512 1536 utf8mb3 utf8mb3_general_ci varchar(512) > +3.0000 information_schema SEQUENCES SEQUENCE_SCHEMA varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +3.0000 information_schema SEQUENCES SEQUENCE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +3.0000 information_schema SEQUENCES DATA_TYPE varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > +NULL information_schema SEQUENCES NUMERIC_PRECISION int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES NUMERIC_PRECISION_RADIX int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES NUMERIC_SCALE int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES START_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES MINIMUM_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES MAXIMUM_VALUE decimal NULL NULL NULL NULL decimal(21,0) > +NULL information_schema SEQUENCES INCREMENT bigint NULL NULL NULL NULL bigint(21) > +NULL information_schema SEQUENCES CYCLE_OPTION bigint NULL NULL NULL NULL bigint(21) > +NULL information_schema SEQUENCES DECLARED_DATA_TYPE int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES DECLARED_NUMERIC_PRECISION int NULL NULL NULL NULL int(21) > +NULL information_schema SEQUENCES DECLARED_NUMERIC_SCALE int NULL NULL NULL NULL int(21) > 3.0000 information_schema SESSION_STATUS VARIABLE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > 3.0000 information_schema SESSION_STATUS VARIABLE_VALUE varchar 2048 6144 utf8mb3 utf8mb3_general_ci varchar(2048) > 3.0000 information_schema SESSION_VARIABLES VARIABLE_NAME varchar 64 192 utf8mb3 utf8mb3_general_ci varchar(64) > diff --git a/mysql-test/suite/funcs_1/r/is_tables_is.result b/mysql-test/suite/funcs_1/r/is_tables_is.result > index c18f733c86f..8c81a79b605 100644 > --- a/mysql-test/suite/funcs_1/r/is_tables_is.result > +++ b/mysql-test/suite/funcs_1/r/is_tables_is.result > @@ -739,6 +739,31 @@ user_comment > Separator ----------------------------------------------------- > TABLE_CATALOG def > TABLE_SCHEMA information_schema > +TABLE_NAME SEQUENCES > +TABLE_TYPE SYSTEM VIEW > +ENGINE MEMORY > +VERSION 11 > +ROW_FORMAT Fixed > +TABLE_ROWS #TBLR# > +AVG_ROW_LENGTH #ARL# > +DATA_LENGTH #DL# > +MAX_DATA_LENGTH #MDL# > +INDEX_LENGTH #IL# > +DATA_FREE #DF# > +AUTO_INCREMENT NULL > +CREATE_TIME #CRT# > +UPDATE_TIME #UT# > +CHECK_TIME #CT# > +TABLE_COLLATION utf8mb3_general_ci > +CHECKSUM NULL > +CREATE_OPTIONS #CO# > +TABLE_COMMENT #TC# > +MAX_INDEX_LENGTH #MIL# > +TEMPORARY Y > +user_comment > +Separator ----------------------------------------------------- > +TABLE_CATALOG def > +TABLE_SCHEMA information_schema > TABLE_NAME SESSION_STATUS > TABLE_TYPE SYSTEM VIEW > ENGINE MEMORY > @@ -1855,6 +1880,31 @@ user_comment > Separator ----------------------------------------------------- > TABLE_CATALOG def > TABLE_SCHEMA information_schema > +TABLE_NAME SEQUENCES > +TABLE_TYPE SYSTEM VIEW > +ENGINE MEMORY > +VERSION 11 > +ROW_FORMAT Fixed > +TABLE_ROWS #TBLR# > +AVG_ROW_LENGTH #ARL# > +DATA_LENGTH #DL# > +MAX_DATA_LENGTH #MDL# > +INDEX_LENGTH #IL# > +DATA_FREE #DF# > +AUTO_INCREMENT NULL > +CREATE_TIME #CRT# > +UPDATE_TIME #UT# > +CHECK_TIME #CT# > +TABLE_COLLATION utf8mb3_general_ci > +CHECKSUM NULL > +CREATE_OPTIONS #CO# > +TABLE_COMMENT #TC# > +MAX_INDEX_LENGTH #MIL# > +TEMPORARY Y > +user_comment > +Separator ----------------------------------------------------- > +TABLE_CATALOG def > +TABLE_SCHEMA information_schema > TABLE_NAME SESSION_STATUS > TABLE_TYPE SYSTEM VIEW > ENGINE MEMORY > diff --git a/mysql-test/suite/funcs_1/r/is_tables_is_embedded.result b/mysql-test/suite/funcs_1/r/is_tables_is_embedded.result > index c18f733c86f..8c81a79b605 100644 > --- a/mysql-test/suite/funcs_1/r/is_tables_is_embedded.result > +++ b/mysql-test/suite/funcs_1/r/is_tables_is_embedded.result > @@ -739,6 +739,31 @@ user_comment > Separator ----------------------------------------------------- > TABLE_CATALOG def > TABLE_SCHEMA information_schema > +TABLE_NAME SEQUENCES > +TABLE_TYPE SYSTEM VIEW > +ENGINE MEMORY > +VERSION 11 > +ROW_FORMAT Fixed > +TABLE_ROWS #TBLR# > +AVG_ROW_LENGTH #ARL# > +DATA_LENGTH #DL# > +MAX_DATA_LENGTH #MDL# > +INDEX_LENGTH #IL# > +DATA_FREE #DF# > +AUTO_INCREMENT NULL > +CREATE_TIME #CRT# > +UPDATE_TIME #UT# > +CHECK_TIME #CT# > +TABLE_COLLATION utf8mb3_general_ci > +CHECKSUM NULL > +CREATE_OPTIONS #CO# > +TABLE_COMMENT #TC# > +MAX_INDEX_LENGTH #MIL# > +TEMPORARY Y > +user_comment > +Separator ----------------------------------------------------- > +TABLE_CATALOG def > +TABLE_SCHEMA information_schema > TABLE_NAME SESSION_STATUS > TABLE_TYPE SYSTEM VIEW > ENGINE MEMORY > @@ -1855,6 +1880,31 @@ user_comment > Separator ----------------------------------------------------- > TABLE_CATALOG def > TABLE_SCHEMA information_schema > +TABLE_NAME SEQUENCES > +TABLE_TYPE SYSTEM VIEW > +ENGINE MEMORY > +VERSION 11 > +ROW_FORMAT Fixed > +TABLE_ROWS #TBLR# > +AVG_ROW_LENGTH #ARL# > +DATA_LENGTH #DL# > +MAX_DATA_LENGTH #MDL# > +INDEX_LENGTH #IL# > +DATA_FREE #DF# > +AUTO_INCREMENT NULL > +CREATE_TIME #CRT# > +UPDATE_TIME #UT# > +CHECK_TIME #CT# > +TABLE_COLLATION utf8mb3_general_ci > +CHECKSUM NULL > +CREATE_OPTIONS #CO# > +TABLE_COMMENT #TC# > +MAX_INDEX_LENGTH #MIL# > +TEMPORARY Y > +user_comment > +Separator ----------------------------------------------------- > +TABLE_CATALOG def > +TABLE_SCHEMA information_schema > TABLE_NAME SESSION_STATUS > TABLE_TYPE SYSTEM VIEW > ENGINE MEMORY > diff --git a/mysql-test/suite/sql_sequence/alter.result b/mysql-test/suite/sql_sequence/alter.result > index 60b708b8289..e6cd599ec40 100644 > --- a/mysql-test/suite/sql_sequence/alter.result > +++ b/mysql-test/suite/sql_sequence/alter.result > @@ -14,7 +14,7 @@ next value for t1 > alter sequence t1 start=50; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 50 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 50 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 2 1 9223372036854775806 50 1 0 0 0 > @@ -24,7 +24,7 @@ next value for t1 > alter sequence t1 minvalue=-100; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 50 minvalue -100 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 50 minvalue -100 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 3 -100 9223372036854775806 50 1 0 0 0 > @@ -33,14 +33,14 @@ ERROR HY000: Sequence 'test.t1' has out of range value for options > alter sequence t1 minvalue=100 start=100 restart=100; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 100 minvalue 100 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 100 minvalue 100 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 100 100 9223372036854775806 100 1 0 0 0 > alter sequence t1 maxvalue=500; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 100 minvalue 100 maxvalue 500 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 100 minvalue 100 maxvalue 500 increment by 1 nocache nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 100 100 500 100 1 0 0 0 > @@ -49,20 +49,20 @@ CREATE SEQUENCE t1 engine=myisam; > alter sequence t1 nocache; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > alter sequence t1 cache=100; > flush tables; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 100 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 100 nocycle ENGINE=MyISAM > alter sequence t1 nocache; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > flush tables; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 nocache nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 0 0 0 > @@ -83,19 +83,19 @@ CREATE SEQUENCE t1 maxvalue=100 engine=myisam; > alter sequence t1 no maxvalue; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 1000 0 0 > alter sequence t1 cycle; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > alter sequence t1 nocycle; > alter sequence t1 start=15 restart minvalue=10 maxvalue=20 cycle; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 15 minvalue 10 maxvalue 20 increment by 1 cache 1000 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 15 minvalue 10 maxvalue 20 increment by 1 cache 1000 cycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 15 10 20 15 1 1000 1 0 > @@ -129,7 +129,7 @@ CREATE SEQUENCE t1 maxvalue=100; > alter sequence t1 increment=-2 start with 50 minvalue=-100; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 50 minvalue -100 maxvalue 100 increment by -2 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 50 minvalue -100 maxvalue 100 increment by -2 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 -100 100 50 -2 1000 0 0 > @@ -159,7 +159,7 @@ next value for t1 > alter sequence t1 start=100; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 100 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 10 nocycle ENGINE=InnoDB > +t1 CREATE SEQUENCE `t1` as bigint start with 100 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 10 nocycle ENGINE=InnoDB > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 11 1 9223372036854775806 100 1 10 0 0 > @@ -185,15 +185,15 @@ next value for t1 > alter table t1 comment="foo"; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB COMMENT='foo' > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB COMMENT='foo' > alter table t1 engine=myisam; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='foo' > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='foo' > alter table t1 engine=innodb; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB COMMENT='foo' > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB COMMENT='foo' > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 3001 1 9223372036854775806 1 1 1000 0 0 > @@ -248,14 +248,186 @@ SELECT NEXTVAL(s); > NEXTVAL(s) > 1 > DROP SEQUENCE s; > +# > +# MDEV-28152 Features for sequence > +# > create sequence s maxvalue 12345; > show create sequence s; > Table Create Table > -s CREATE SEQUENCE `s` start with 1 minvalue 1 maxvalue 12345 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 12345 increment by 1 cache 1000 nocycle ENGINE=MyISAM > alter sequence s maxvalue 123456789012345678901234; > Warnings: > Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > show create sequence s; > Table Create Table > -s CREATE SEQUENCE `s` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +drop sequence s; > +create sequence s as tinyint; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as tinyint start with 1 minvalue 1 maxvalue 126 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` tinyint(5) NOT NULL, > + `minimum_value` tinyint(5) NOT NULL, > + `maximum_value` tinyint(5) NOT NULL, > + `start_value` tinyint(5) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s as int; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as int start with 1 minvalue 1 maxvalue 126 increment by 1 cache 1000 nocycle ENGINE=MyISAM Here (and for each type) this should be tested: 1. check maxvalue cycling and reaching maximum in case of not cycling which was inherited (126 in the case) 2. alter maxvalue according the new type 2.1 make it higher then possible (get a error or fixing the value with a warning) 2.2 make it maximum allowed and test it works 2.3. make it lower (may be around middle range) and test if it works > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` int(12) NOT NULL, > + `minimum_value` int(12) NOT NULL, > + `maximum_value` int(12) NOT NULL, > + `start_value` int(12) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s maxvalue 12345; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as int start with 1 minvalue 1 maxvalue 12345 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` int(12) NOT NULL, > + `minimum_value` int(12) NOT NULL, > + `maximum_value` int(12) NOT NULL, > + `start_value` int(12) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +drop sequence s; > +create sequence s; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` bigint(21) NOT NULL, > + `minimum_value` bigint(21) NOT NULL, > + `maximum_value` bigint(21) NOT NULL, > + `start_value` bigint(21) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s maxvalue 123; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 123 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` bigint(21) NOT NULL, > + `minimum_value` bigint(21) NOT NULL, > + `maximum_value` bigint(21) NOT NULL, > + `start_value` bigint(21) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s as tinyint; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as tinyint start with 1 minvalue 1 maxvalue 123 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` tinyint(5) NOT NULL, > + `minimum_value` tinyint(5) NOT NULL, > + `maximum_value` tinyint(5) NOT NULL, > + `start_value` tinyint(5) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +drop sequence s; > +create sequence s as int; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as int start with 1 minvalue 1 maxvalue 2147483646 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` int(12) NOT NULL, > + `minimum_value` int(12) NOT NULL, > + `maximum_value` int(12) NOT NULL, > + `start_value` int(12) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s as tinyint; > +ERROR 22003: Out of range value for column 'maximum_value' at row 1 > +drop sequence s; Here it would be nice (before that drop) to change maximum_value and then repeat operation to be sure it is allowed. alter sequence s maxvalue 126; alter sequence s as tinyint; > +create sequence s as tinyint; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as tinyint start with 1 minvalue 1 maxvalue 126 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` tinyint(5) NOT NULL, > + `minimum_value` tinyint(5) NOT NULL, > + `maximum_value` tinyint(5) NOT NULL, > + `start_value` tinyint(5) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s as int maxvalue 123; > +ERROR 42000: This version of MariaDB doesn't yet support 'ALTER SEQUENCE with both AS <type> and something else.' > +drop sequence s; > +create sequence s as int; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as int start with 1 minvalue 1 maxvalue 2147483646 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` int(12) NOT NULL, > + `minimum_value` int(12) NOT NULL, > + `maximum_value` int(12) NOT NULL, > + `start_value` int(12) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +alter sequence s as int unsigned; > +show create sequence s; > +Table Create Table > +s CREATE SEQUENCE `s` as int unsigned start with 1 minvalue 1 maxvalue 2147483646 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table s; > +Table Create Table > +s CREATE TABLE `s` ( > + `next_not_cached_value` int(12) unsigned NOT NULL, > + `minimum_value` int(12) unsigned NOT NULL, > + `maximum_value` int(12) unsigned NOT NULL, > + `start_value` int(12) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > drop sequence s; > diff --git a/mysql-test/suite/sql_sequence/alter.test b/mysql-test/suite/sql_sequence/alter.test > index 3afad38a9e5..4949075f492 100644 > --- a/mysql-test/suite/sql_sequence/alter.test > +++ b/mysql-test/suite/sql_sequence/alter.test > @@ -162,12 +162,63 @@ ALTER TABLE s ORDER BY cache_size; > SELECT NEXTVAL(s); > DROP SEQUENCE s; > > -# > -# MDEV-28152 Features for sequence > -# > > +--echo # > +--echo # MDEV-28152 Features for sequence > +--echo # > + > +# truncation in alter sequence > create sequence s maxvalue 12345; > show create sequence s; > alter sequence s maxvalue 123456789012345678901234; > show create sequence s; > drop sequence s; > + > +# alter first from a narrower type to a wider type, then maxvalue > +create sequence s as tinyint; > +show create sequence s; > +show create table s; > +alter sequence s as int; > +show create sequence s; > +show create table s; > +alter sequence s maxvalue 12345; > +show create sequence s; > +show create table s; > +drop sequence s; > + > +# alter first maxvalue then from a wider type to a narrower type > +create sequence s; > +show create sequence s; > +show create table s; > +alter sequence s maxvalue 123; > +show create sequence s; > +show create table s; > +alter sequence s as tinyint; > +show create sequence s; > +show create table s; > +drop sequence s; > + > +# from a wider type to a narrower type with out of range values > +create sequence s as int; > +show create sequence s; > +show create table s; > +--error ER_WARN_DATA_OUT_OF_RANGE > +alter sequence s as tinyint; > +drop sequence s; > + > +# cannot alter both value type and something else yet. > +create sequence s as tinyint; > +show create sequence s; > +show create table s; > +--error ER_NOT_SUPPORTED_YET > +alter sequence s as int maxvalue 123; > +drop sequence s; > + > +# from signed to unsigned > +create sequence s as int; > +show create sequence s; > +show create table s; > +alter sequence s as int unsigned; > +show create sequence s; > +show create table s; > +drop sequence s; > diff --git a/mysql-test/suite/sql_sequence/alter_notembedded.result b/mysql-test/suite/sql_sequence/alter_notembedded.result > index f3e1f5f18cd..c2ea6633d1d 100644 > --- a/mysql-test/suite/sql_sequence/alter_notembedded.result > +++ b/mysql-test/suite/sql_sequence/alter_notembedded.result > @@ -12,7 +12,7 @@ select nextval(s1); > ERROR 42000: INSERT command denied to user 'normal_1'@'localhost' for table `s_db`.`s1` > show create sequence s1; > Table Create Table > -s1 CREATE SEQUENCE `s1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +s1 CREATE SEQUENCE `s1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > alter sequence s1 restart 50; > ERROR 42000: ALTER command denied to user 'normal_1'@'localhost' for table `s_db`.`s1` > connection default; > diff --git a/mysql-test/suite/sql_sequence/aria.result b/mysql-test/suite/sql_sequence/aria.result > index cfc7d946772..1e42be58240 100644 > --- a/mysql-test/suite/sql_sequence/aria.result > +++ b/mysql-test/suite/sql_sequence/aria.result > @@ -2,7 +2,7 @@ set @@default_storage_engine="aria"; > CREATE SEQUENCE t1 cache=10; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 10 nocycle ENGINE=Aria > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 10 nocycle ENGINE=Aria > select NEXT VALUE for t1,seq from seq_1_to_20; > NEXT VALUE for t1 seq > 1 1 > diff --git a/mysql-test/suite/sql_sequence/concurrent_create.result b/mysql-test/suite/sql_sequence/concurrent_create.result > index 2473abef37d..e28c98e46be 100644 > --- a/mysql-test/suite/sql_sequence/concurrent_create.result > +++ b/mysql-test/suite/sql_sequence/concurrent_create.result > @@ -22,6 +22,20 @@ select * from s2; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 1000 0 0 > DROP SEQUENCE s1, s2; > +# > +# MDEV-28152 Features for sequence > +# > +CREATE SEQUENCE s1 as mediumint unsigned ENGINE=InnoDB; > +PREPARE stmt FROM "CREATE TABLE s2 LIKE s1"; > +execute stmt; > +drop table s2; > +execute stmt; > +drop table s2; > +execute stmt; > +select * from s2; > +next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > +1 1 16777214 1 1 1000 0 0 > +DROP SEQUENCE s1, s2; > CREATE SEQUENCE s1 ENGINE=InnoDB; > connect con1,localhost,root,,test; > CREATE TABLE s2 LIKE s1;; > diff --git a/mysql-test/suite/sql_sequence/concurrent_create.test b/mysql-test/suite/sql_sequence/concurrent_create.test > index b27a6d3bdb9..b4def8d7ce5 100644 > --- a/mysql-test/suite/sql_sequence/concurrent_create.test > +++ b/mysql-test/suite/sql_sequence/concurrent_create.test > @@ -38,6 +38,20 @@ execute stmt; > select * from s2; > DROP SEQUENCE s1, s2; > > +--echo # > +--echo # MDEV-28152 Features for sequence > +--echo # > + > +CREATE SEQUENCE s1 as mediumint unsigned ENGINE=InnoDB; > +PREPARE stmt FROM "CREATE TABLE s2 LIKE s1"; > +execute stmt; > +drop table s2; > +execute stmt; > +drop table s2; > +execute stmt; > +select * from s2; > +DROP SEQUENCE s1, s2; You also better check second execution of CREATE SEQUENCE with the type in SP/PS (probably not in this test file) > + > # > # MDEV-15117 Server crashes in in open_and_process_table or ASAN > # heap-use-after-free in is_temporary_table upon creating/flushing sequences > diff --git a/mysql-test/suite/sql_sequence/create.result b/mysql-test/suite/sql_sequence/create.result > index e6a382ec3bf..f2d376774a7 100644 > --- a/mysql-test/suite/sql_sequence/create.result > +++ b/mysql-test/suite/sql_sequence/create.result > @@ -4,7 +4,7 @@ Note 1051 Unknown table 'test.t1' > create or replace sequence t1 engine=myisam; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -23,7 +23,7 @@ next_not_cached_value minimum_value maximum_value start_value increment cache_si > create or replace sequence t1 engine=innodb; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -42,7 +42,7 @@ next_not_cached_value minimum_value maximum_value start_value increment cache_si > create or replace sequence t1 engine=maria; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -65,59 +65,59 @@ ERROR 42S02: Table 'test.t1' doesn't exist > create or replace sequence t1 start with 10; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 10 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 10 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 10 1 9223372036854775806 10 1 1000 0 0 > create or replace sequence t1 minvalue=11; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 11 minvalue 11 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 11 minvalue 11 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 11 11 9223372036854775806 11 1 1000 0 0 > create or replace sequence t1 maxvalue=13 increment by -1; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 13 minvalue -9223372036854775807 maxvalue 13 increment by -1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 13 minvalue -9223372036854775807 maxvalue 13 increment by -1 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 13 -9223372036854775807 13 13 -1 1000 0 0 > create or replace sequence t1 increment by -1 cache 100; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with -1 minvalue -9223372036854775807 maxvalue -1 increment by -1 cache 100 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with -1 minvalue -9223372036854775807 maxvalue -1 increment by -1 cache 100 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > -1 -9223372036854775807 -1 -1 -1 100 0 0 > create or replace sequence t1 cycle; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 1000 1 0 > create or replace sequence t1 nocycle; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 1000 0 0 > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > create or replace sequence t1 cycle minvalue= 14; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 14 minvalue 14 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 14 minvalue 14 maxvalue 9223372036854775806 increment by 1 cache 1000 cycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 14 14 9223372036854775806 14 1 1000 1 0 > create or replace sequence t1 cycle increment by -1; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with -1 minvalue -9223372036854775807 maxvalue -1 increment by -1 cache 1000 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with -1 minvalue -9223372036854775807 maxvalue -1 increment by -1 cache 1000 cycle ENGINE=MyISAM > drop sequence t1; > create sequence if not exists t1; > create sequence if not exists t1 start with 10; > @@ -128,81 +128,21 @@ next_not_cached_value minimum_value maximum_value start_value increment cache_si > 1 1 9223372036854775806 1 1 1000 0 0 > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM For backward compatibility (it is present in mariadbdump) and reducing sizeof the patch we can not print "as bigint" if it is bigint, because it is default. > create or replace sequence t1 start with 10 minvalue=10 maxvalue=11 nocache cycle; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 10 minvalue 10 maxvalue 11 increment by 1 nocache cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 10 minvalue 10 maxvalue 11 increment by 1 nocache cycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 10 10 11 10 1 0 1 0 > create or replace sequence t1 start with 10 minvalue=-10 maxvalue=11 cache=10 cycle increment by 10; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 10 minvalue -10 maxvalue 11 increment by 10 cache 10 cycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 10 minvalue -10 maxvalue 11 increment by 10 cache 10 cycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 10 -10 11 10 10 10 1 0 > -create or replace sequence t1 minvalue -999999999999999999999; > -Warnings: > -Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 minvalue -9223372036854775808; > -Warnings: > -Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 minvalue -9223372036854775807; > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 minvalue 9223372036854775805; > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with 9223372036854775805 minvalue 9223372036854775805 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 minvalue 9223372036854775806; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 minvalue 9223372036854775807; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 minvalue 9223372036854775808; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 minvalue 9999999999999999999999; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 maxvalue -999999999999999999999 increment by -1; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 maxvalue -9223372036854775808 increment by -1; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 maxvalue -9223372036854775807 increment by -1; > -ERROR HY000: Sequence 'test.t1' has out of range value for options > -create or replace sequence t1 maxvalue -9223372036854775806 increment by -1; > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with -9223372036854775806 minvalue -9223372036854775807 maxvalue -9223372036854775806 increment by -1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 maxvalue 9223372036854775806; > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 maxvalue 9223372036854775807; > -Warnings: > -Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 maxvalue 9223372036854775808; > -Warnings: > -Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > -create or replace sequence t1 maxvalue 9999999999999999999999; > -Warnings: > -Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > -show create sequence t1; > -Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM Why was it deleted? > create or replace sequence t1 start with 10 NO MAXVALUE NO MINVALUE; > create or replace sequence t1 start with 10 maxvalue 10; > create or replace sequence t1 start with 10 minvalue 10; > @@ -213,7 +153,7 @@ drop sequence if exists t1; > create sequence t1 increment by 0; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 0 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 0 cache 1000 nocycle ENGINE=MyISAM > select * from t1; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 0 1000 0 0 > @@ -293,7 +233,7 @@ flush tables; > create or replace sequence t1 comment= "test 1"; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='test 1' > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='test 1' > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -309,7 +249,7 @@ t1 CREATE TABLE `t1` ( > create or replace sequence t1 comment= "test 2" min_rows=1 max_rows=2; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='test 2' > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM COMMENT='test 2' > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -338,7 +278,7 @@ CREATE TABLE t1 ( > ) sequence=1; > show create sequence t1; > Table Create Table > -t1 CREATE SEQUENCE `t1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > show create table t1; > Table Create Table > t1 CREATE TABLE `t1` ( > @@ -373,7 +313,7 @@ CREATE OR REPLACE TABLE t1 ( > `cycle_option` tinyint(1) unsigned NOT NULL, > `cycle_count` bigint(21) NOT NULL > ) sequence=1; > -ERROR HY000: Sequence 'test.t1' table structure is invalid (next_not_cached_value) > +ERROR HY000: Sequence 'test.t1' table structure is invalid (minimum_value) > CREATE OR REPLACE TABLE t1 ( > `next_not_cached_value` bigint(21) NOT NULL, > `minimum_value` bigint(21) NOT NULL, > @@ -607,7 +547,7 @@ t CREATE TABLE `t` ( > ) ENGINE=MyISAM SEQUENCE=1 > show create sequence t; > Table Create Table > -t CREATE SEQUENCE `t` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +t CREATE SEQUENCE `t` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > drop tables t, s; > # > # MDEV-13714 SEQUENCE option fix > @@ -772,3 +712,439 @@ CREATE TRIGGER s1 BEFORE UPDATE ON seq1 FOR EACH ROW SET @a= 5; > ERROR HY000: Trigger's 'seq1' is a view, temporary table or sequence > DROP SEQUENCE seq1; > # End of 10.4 test > +###### > +# MDEV-28152 Features for sequence > +###### > +# ----- > +# Truncating out-of-bound numbers for minvalue and maxvalue > +# ----- > +create or replace sequence t1 minvalue -999999999999999999999; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 minvalue -9223372036854775808; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 minvalue -9223372036854775807; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with -9223372036854775807 minvalue -9223372036854775807 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 minvalue 9223372036854775805; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 9223372036854775805 minvalue 9223372036854775805 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 minvalue 9223372036854775806; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 minvalue 9223372036854775807; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 minvalue 9223372036854775808; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 minvalue 9999999999999999999999; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 maxvalue -999999999999999999999 increment by -1; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 maxvalue -9223372036854775808 increment by -1; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 maxvalue -9223372036854775807 increment by -1; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 maxvalue -9223372036854775806 increment by -1; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with -9223372036854775806 minvalue -9223372036854775807 maxvalue -9223372036854775806 increment by -1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 maxvalue 9223372036854775806; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 maxvalue 9223372036854775807; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 maxvalue 9223372036854775808; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 maxvalue 9999999999999999999999; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +# ----- > +# Create with value types > +# ----- > +create or replace sequence t1 as tinyint; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tinyint start with 1 minvalue 1 maxvalue 126 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) NOT NULL, > + `minimum_value` tinyint(5) NOT NULL, > + `maximum_value` tinyint(5) NOT NULL, > + `start_value` tinyint(5) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as smallint; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as smallint start with 1 minvalue 1 maxvalue 32766 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` smallint(7) NOT NULL, > + `minimum_value` smallint(7) NOT NULL, > + `maximum_value` smallint(7) NOT NULL, > + `start_value` smallint(7) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as mediumint; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as mediumint start with 1 minvalue 1 maxvalue 8388606 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` mediumint(10) NOT NULL, > + `minimum_value` mediumint(10) NOT NULL, > + `maximum_value` mediumint(10) NOT NULL, > + `start_value` mediumint(10) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as int; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as int start with 1 minvalue 1 maxvalue 2147483646 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` int(12) NOT NULL, > + `minimum_value` int(12) NOT NULL, > + `maximum_value` int(12) NOT NULL, > + `start_value` int(12) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as bigint; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) NOT NULL, > + `minimum_value` bigint(21) NOT NULL, > + `maximum_value` bigint(21) NOT NULL, > + `start_value` bigint(21) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as tinyint unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tiny unsigned start with 1 minvalue 1 maxvalue 254 increment by 1 cache 1000 nocycle ENGINE=MyISAM You better also check the minimum possible value for each signed and unsigned type (I see default 1 in most cases here, it would be nice to see legal minimums checked (illegal as I can see tested later)). > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) unsigned NOT NULL, > + `minimum_value` tinyint(5) unsigned NOT NULL, > + `maximum_value` tinyint(5) unsigned NOT NULL, > + `start_value` tinyint(5) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as smallint unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as smallint unsigned start with 1 minvalue 1 maxvalue 65534 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` smallint(7) unsigned NOT NULL, > + `minimum_value` smallint(7) unsigned NOT NULL, > + `maximum_value` smallint(7) unsigned NOT NULL, > + `start_value` smallint(7) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as mediumint unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as mediumint unsigned start with 1 minvalue 1 maxvalue 16777214 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` mediumint(10) unsigned NOT NULL, > + `minimum_value` mediumint(10) unsigned NOT NULL, > + `maximum_value` mediumint(10) unsigned NOT NULL, > + `start_value` mediumint(10) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as int unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as int unsigned start with 1 minvalue 1 maxvalue 4294967294 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` int(12) unsigned NOT NULL, > + `minimum_value` int(12) unsigned NOT NULL, > + `maximum_value` int(12) unsigned NOT NULL, > + `start_value` int(12) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as bigint unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint unsigned start with 1 minvalue 1 maxvalue 18446744073709551614 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) unsigned NOT NULL, > + `minimum_value` bigint(21) unsigned NOT NULL, > + `maximum_value` bigint(21) unsigned NOT NULL, > + `start_value` bigint(21) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as tinyint zerofill; > +ERROR HY000: Incorrect value 'ZEROFILL' for option 'AS' > +create or replace sequence t1 as mediumint unsigned increment by -1; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 as bigint unsigned start with 12345678901234567890; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint unsigned start with 12345678901234567890 minvalue 1 maxvalue 18446744073709551614 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) unsigned NOT NULL, > + `minimum_value` bigint(21) unsigned NOT NULL, > + `maximum_value` bigint(21) unsigned NOT NULL, > + `start_value` bigint(21) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +# ----- > +# value types + truncating > +# ----- > +create or replace sequence t1 minvalue -23 maxvalue 9999 as tinyint unsigned; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tiny unsigned start with 1 minvalue 1 maxvalue 254 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) unsigned NOT NULL, > + `minimum_value` tinyint(5) unsigned NOT NULL, > + `maximum_value` tinyint(5) unsigned NOT NULL, > + `start_value` tinyint(5) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 minvalue -32768 maxvalue 32767 as smallint; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as smallint start with -32767 minvalue -32767 maxvalue 32766 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` smallint(7) NOT NULL, > + `minimum_value` smallint(7) NOT NULL, > + `maximum_value` smallint(7) NOT NULL, > + `start_value` smallint(7) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 minvalue 0 maxvalue 65535 as smallint unsigned; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as smallint unsigned start with 1 minvalue 1 maxvalue 65534 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` smallint(7) unsigned NOT NULL, > + `minimum_value` smallint(7) unsigned NOT NULL, > + `maximum_value` smallint(7) unsigned NOT NULL, > + `start_value` smallint(7) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 minvalue -12345678901234 as mediumint unsigned maxvalue 12345678901234; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as mediumint unsigned start with 1 minvalue 1 maxvalue 16777214 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` mediumint(10) unsigned NOT NULL, > + `minimum_value` mediumint(10) unsigned NOT NULL, > + `maximum_value` mediumint(10) unsigned NOT NULL, > + `start_value` mediumint(10) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as bigint unsigned minvalue -12345678901234 maxvalue 12345678901234; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MINVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint unsigned start with 1 minvalue 1 maxvalue 12345678901234 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) unsigned NOT NULL, > + `minimum_value` bigint(21) unsigned NOT NULL, > + `maximum_value` bigint(21) unsigned NOT NULL, > + `start_value` bigint(21) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +# ----- > +# indistinguishable values during parsing if we did not pass back Longlong_hybrid from the parser. > +# ----- > +create or replace sequence t1 as bigint maxvalue -1 increment by -1; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with -1 minvalue -9223372036854775807 maxvalue -1 increment by -1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) NOT NULL, > + `minimum_value` bigint(21) NOT NULL, > + `maximum_value` bigint(21) NOT NULL, > + `start_value` bigint(21) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as bigint maxvalue 18446744073709551615; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) NOT NULL, > + `minimum_value` bigint(21) NOT NULL, > + `maximum_value` bigint(21) NOT NULL, > + `start_value` bigint(21) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +create or replace sequence t1 as bigint unsigned maxvalue -1; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 as bigint unsigned maxvalue 18446744073709551615; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint unsigned start with 1 minvalue 1 maxvalue 18446744073709551614 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` bigint(21) unsigned NOT NULL, > + `minimum_value` bigint(21) unsigned NOT NULL, > + `maximum_value` bigint(21) unsigned NOT NULL, > + `start_value` bigint(21) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +# ----- > +# value types + out of range start > +# ----- > +create or replace sequence t1 start with -123456789012345678901 as tinyint unsigned; > +ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near '123456789012345678901 as tinyint unsigned' at line 1 > +create or replace sequence t1 start with -1 as tinyint unsigned; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 start with 0 as tinyint unsigned; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as bigint unsigned start with 1 minvalue 1 maxvalue 18446744073709551614 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 start with 1 as tinyint unsigned; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tiny unsigned start with 1 minvalue 1 maxvalue 254 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +create or replace sequence t1 start with 254 as tinyint unsigned; > +create or replace sequence t1 start with 255 as tinyint unsigned; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 start with 256 as tinyint unsigned; > +ERROR HY000: Sequence 'test.t1' has out of range value for options > +create or replace sequence t1 start with 123456789012345678901 as tinyint unsigned; > +ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near '123456789012345678901 as tinyint unsigned' at line 1 > +drop sequence t1; > +create sequence s1 as tinyint unsigned increment by 23; > +create sequence s2 start with 42 minvalue -9223372036854775807; > +create sequence s3 as bigint unsigned start with 12345678901234567890 cycle; > +select * from information_schema.sequences; > +SEQUENCE_CATALOG SEQUENCE_SCHEMA SEQUENCE_NAME DATA_TYPE NUMERIC_PRECISION NUMERIC_PRECISION_RADIX NUMERIC_SCALE START_VALUE MINIMUM_VALUE MAXIMUM_VALUE INCREMENT CYCLE_OPTION DECLARED_DATA_TYPE DECLARED_NUMERIC_PRECISION DECLARED_NUMERIC_SCALE > +def test s3 bigint unsigned 64 2 0 12345678901234567890 1 18446744073709551614 1 1 NULL NULL NULL > +def test s2 bigint 64 2 0 42 -9223372036854775807 9223372036854775806 1 0 NULL NULL NULL > +def test s1 tiny unsigned 8 2 0 1 1 254 23 0 NULL NULL NULL > +drop sequence s1, s2, s3; > diff --git a/mysql-test/suite/sql_sequence/create.test b/mysql-test/suite/sql_sequence/create.test > index 54e181483d6..388b2fd3fe9 100644 > --- a/mysql-test/suite/sql_sequence/create.test > +++ b/mysql-test/suite/sql_sequence/create.test > @@ -71,41 +71,6 @@ create or replace sequence t1 start with 10 minvalue=-10 maxvalue=11 cache=10 cy > show create sequence t1; > select * from t1; > > -# Truncating out-of-bound numbers for minvalue and maxvalue > -create or replace sequence t1 minvalue -999999999999999999999; > -show create sequence t1; > -create or replace sequence t1 minvalue -9223372036854775808; > -show create sequence t1; > -create or replace sequence t1 minvalue -9223372036854775807; > -show create sequence t1; > -create or replace sequence t1 minvalue 9223372036854775805; > -show create sequence t1; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 minvalue 9223372036854775806; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 minvalue 9223372036854775807; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 minvalue 9223372036854775808; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 minvalue 9999999999999999999999; > - > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 maxvalue -999999999999999999999 increment by -1; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 maxvalue -9223372036854775808 increment by -1; > ---error ER_SEQUENCE_INVALID_DATA > -create or replace sequence t1 maxvalue -9223372036854775807 increment by -1; > -create or replace sequence t1 maxvalue -9223372036854775806 increment by -1; > -show create sequence t1; > -create or replace sequence t1 maxvalue 9223372036854775806; > -show create sequence t1; > -create or replace sequence t1 maxvalue 9223372036854775807; > -show create sequence t1; > -create or replace sequence t1 maxvalue 9223372036854775808; > -show create sequence t1; > -create or replace sequence t1 maxvalue 9999999999999999999999; > -show create sequence t1; > - > # NO MINVALUE, NO MAXVALUE > create or replace sequence t1 start with 10 NO MAXVALUE NO MINVALUE; > > @@ -582,3 +547,160 @@ CREATE TRIGGER s1 BEFORE UPDATE ON seq1 FOR EACH ROW SET @a= 5; > DROP SEQUENCE seq1; > > --echo # End of 10.4 test > + > +--echo ###### > +--echo # MDEV-28152 Features for sequence > +--echo ###### > + > +--echo # ----- > +--echo # Truncating out-of-bound numbers for minvalue and maxvalue > +--echo # ----- > +--disable_ps_protocol > +create or replace sequence t1 minvalue -999999999999999999999; > +show create sequence t1; > +create or replace sequence t1 minvalue -9223372036854775808; > +show create sequence t1; > +create or replace sequence t1 minvalue -9223372036854775807; > +show create sequence t1; > +create or replace sequence t1 minvalue 9223372036854775805; > +show create sequence t1; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 minvalue 9223372036854775806; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 minvalue 9223372036854775807; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 minvalue 9223372036854775808; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 minvalue 9999999999999999999999; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 maxvalue -999999999999999999999 increment by -1; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 maxvalue -9223372036854775808 increment by -1; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 maxvalue -9223372036854775807 increment by -1; > +create or replace sequence t1 maxvalue -9223372036854775806 increment by -1; > +show create sequence t1; > +create or replace sequence t1 maxvalue 9223372036854775806; > +show create sequence t1; > +create or replace sequence t1 maxvalue 9223372036854775807; > +show create sequence t1; > +create or replace sequence t1 maxvalue 9223372036854775808; > +show create sequence t1; > +create or replace sequence t1 maxvalue 9999999999999999999999; > +show create sequence t1; > +--enable_ps_protocol > + > +--echo # ----- > +--echo # Create with value types > +--echo # ----- > +create or replace sequence t1 as tinyint; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as smallint; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as mediumint; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as int; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as bigint; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as tinyint unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as smallint unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as mediumint unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as int unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as bigint unsigned; > +show create sequence t1; > +show create table t1; > +#zerofill is not supported > +--error ER_BAD_OPTION_VALUE > +create or replace sequence t1 as tinyint zerofill; > +#an unsigned sequence has to have positive increment > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 as mediumint unsigned increment by -1; > +#start with a number between longlong_max and ulonglong_max > +create or replace sequence t1 as bigint unsigned start with 12345678901234567890; > +show create sequence t1; > +show create table t1; > + > +--echo # ----- > +--echo # value types + truncating > +--echo # ----- > +--disable_ps_protocol > +create or replace sequence t1 minvalue -23 maxvalue 9999 as tinyint unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 minvalue -32768 maxvalue 32767 as smallint; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 minvalue 0 maxvalue 65535 as smallint unsigned; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 minvalue -12345678901234 as mediumint unsigned maxvalue 12345678901234; > +show create sequence t1; > +show create table t1; > +create or replace sequence t1 as bigint unsigned minvalue -12345678901234 maxvalue 12345678901234; > +show create sequence t1; > +show create table t1; > +--enable_ps_protocol > + > +--echo # ----- > +--echo # indistinguishable values during parsing if we did not pass back Longlong_hybrid from the parser. > +--echo # ----- > +#signed, -1: no truncation. Note that we need a negative increment because this is a nagative sequence > +create or replace sequence t1 as bigint maxvalue -1 increment by -1; > +show create sequence t1; > +show create table t1; > +--disable_ps_protocol > +#signed, ulonglong_max: turncating to longlong_max-1 > +create or replace sequence t1 as bigint maxvalue 18446744073709551615; > +show create sequence t1; > +show create table t1; > +#unsigned, -1: truncation and invalid data (max_value truncated to 1 which is equal to min_value) > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 as bigint unsigned maxvalue -1; > +#unsigned, ulonglong_max: truncating to ulonglong_max-1 > +create or replace sequence t1 as bigint unsigned maxvalue 18446744073709551615; > +show create sequence t1; > +show create table t1; > +--enable_ps_protocol > + > +--echo # ----- > +--echo # value types + out of range start > +--echo # ----- > +--error ER_PARSE_ERROR > +create or replace sequence t1 start with -123456789012345678901 as tinyint unsigned; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 start with -1 as tinyint unsigned; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 start with 0 as tinyint unsigned; > +show create sequence t1; > +create or replace sequence t1 start with 1 as tinyint unsigned; > +show create sequence t1; > +create or replace sequence t1 start with 254 as tinyint unsigned; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 start with 255 as tinyint unsigned; > +--error ER_SEQUENCE_INVALID_DATA > +create or replace sequence t1 start with 256 as tinyint unsigned; > +--error ER_PARSE_ERROR > +create or replace sequence t1 start with 123456789012345678901 as tinyint unsigned; > + > +drop sequence t1; > + > +# information_schema.sequences > +create sequence s1 as tinyint unsigned increment by 23; > +create sequence s2 start with 42 minvalue -9223372036854775807; > +create sequence s3 as bigint unsigned start with 12345678901234567890 cycle; > +select * from information_schema.sequences; > +drop sequence s1, s2, s3; > diff --git a/mysql-test/suite/sql_sequence/mysqldump.result b/mysql-test/suite/sql_sequence/mysqldump.result > index 0199bb7162f..3316fdac5ae 100644 > --- a/mysql-test/suite/sql_sequence/mysqldump.result > +++ b/mysql-test/suite/sql_sequence/mysqldump.result > @@ -3,9 +3,9 @@ CREATE TABLE t1(a INT, KEY (a)) KEY_BLOCK_SIZE=1024; > insert into t1 values (1),(2); > CREATE SEQUENCE x1 engine=innodb; > # dump whole database > -CREATE SEQUENCE `a1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > +CREATE SEQUENCE `a1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > SELECT SETVAL(`a1`, 1, 0); > -CREATE SEQUENCE `x1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > +CREATE SEQUENCE `x1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > SELECT SETVAL(`x1`, 1, 0); > /*!40101 SET @saved_cs_client = @@character_set_client */; > /*!40101 SET character_set_client = utf8 */; > @@ -18,9 +18,9 @@ INSERT INTO `t1` VALUES > (1), > (2); > # dump by tables order 1 > -CREATE SEQUENCE `a1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > +CREATE SEQUENCE `a1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > SELECT SETVAL(`a1`, 1, 0); > -CREATE SEQUENCE `x1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > +CREATE SEQUENCE `x1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > SELECT SETVAL(`x1`, 1, 0); > /*!40101 SET @saved_cs_client = @@character_set_client */; > /*!40101 SET character_set_client = utf8 */; > @@ -33,9 +33,9 @@ INSERT INTO `t1` VALUES > (1), > (2); > # dump by tables order 2 > -CREATE SEQUENCE `a1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > +CREATE SEQUENCE `a1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > SELECT SETVAL(`a1`, 1, 0); > -CREATE SEQUENCE `x1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > +CREATE SEQUENCE `x1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > SELECT SETVAL(`x1`, 1, 0); > /*!40101 SET @saved_cs_client = @@character_set_client */; > /*!40101 SET character_set_client = utf8 */; > @@ -59,9 +59,9 @@ INSERT INTO `t1` VALUES > (1), > (2); > # dump by tables only sequences > -CREATE SEQUENCE `a1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > +CREATE SEQUENCE `a1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=Aria; > SELECT SETVAL(`a1`, 1, 0); > -CREATE SEQUENCE `x1` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > +CREATE SEQUENCE `x1` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=InnoDB; > SELECT SETVAL(`x1`, 1, 0); > # end of dumps > DROP TABLE a1,t1,x1; > diff --git a/mysql-test/suite/sql_sequence/next.result b/mysql-test/suite/sql_sequence/next.result > index 9d55921006b..76f42143305 100644 > --- a/mysql-test/suite/sql_sequence/next.result > +++ b/mysql-test/suite/sql_sequence/next.result > @@ -548,3 +548,92 @@ SELECT SETVAL (v,0); > ERROR 42S02: 'test.v' is not a SEQUENCE > UNLOCK TABLES; > DROP VIEW v; > +# > +# MDEV-28152 Features for sequence > +# > +create or replace sequence t1 as tinyint unsigned minvalue 1 maxvalue 2; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tiny unsigned start with 1 minvalue 1 maxvalue 2 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) unsigned NOT NULL, > + `minimum_value` tinyint(5) unsigned NOT NULL, > + `maximum_value` tinyint(5) unsigned NOT NULL, > + `start_value` tinyint(5) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +select * from t1; > +next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > +1 1 2 1 1 1000 0 0 > +select next value for t1; > +next value for t1 > +1 > +select next value for t1; > +next value for t1 > +2 > +select next value for t1; > +ERROR HY000: Sequence 'test.t1' has run out > +create or replace sequence t1 as tinyint unsigned minvalue 1 maxvalue 2 cycle; > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tiny unsigned start with 1 minvalue 1 maxvalue 2 increment by 1 cache 1000 cycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) unsigned NOT NULL, > + `minimum_value` tinyint(5) unsigned NOT NULL, > + `maximum_value` tinyint(5) unsigned NOT NULL, > + `start_value` tinyint(5) unsigned NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +select * from t1; > +next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > +1 1 2 1 1 1000 1 0 > +select next value for t1; > +next value for t1 > +1 > +select next value for t1; > +next value for t1 > +2 > +select next value for t1; > +next value for t1 > +1 > +create or replace sequence t1 minvalue -23 maxvalue 99999 as tinyint; > +Warnings: > +Note 1292 Truncated incorrect INTEGER value: 'MAXVALUE' > +show create sequence t1; > +Table Create Table > +t1 CREATE SEQUENCE `t1` as tinyint start with -23 minvalue -23 maxvalue 126 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +show create table t1; > +Table Create Table > +t1 CREATE TABLE `t1` ( > + `next_not_cached_value` tinyint(5) NOT NULL, > + `minimum_value` tinyint(5) NOT NULL, > + `maximum_value` tinyint(5) NOT NULL, > + `start_value` tinyint(5) NOT NULL COMMENT 'start value when sequences is created or value if RESTART is used', > + `increment` bigint(21) NOT NULL COMMENT 'increment value', > + `cache_size` bigint(21) unsigned NOT NULL, > + `cycle_option` tinyint(1) unsigned NOT NULL COMMENT '0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed', > + `cycle_count` bigint(21) NOT NULL COMMENT 'How many cycles have been done' > +) ENGINE=MyISAM SEQUENCE=1 > +select * from t1; > +next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > +-23 -23 126 -23 1 1000 0 0 > +select next value for t1; > +next value for t1 > +-23 > +select next value for t1; > +next value for t1 > +-22 > +select next value for t1; > +next value for t1 > +-21 > +drop sequence t1; > diff --git a/mysql-test/suite/sql_sequence/next.test b/mysql-test/suite/sql_sequence/next.test > index a80f9fad561..f2054e5f116 100644 > --- a/mysql-test/suite/sql_sequence/next.test > +++ b/mysql-test/suite/sql_sequence/next.test > @@ -297,3 +297,36 @@ SELECT SETVAL (v,0); > > UNLOCK TABLES; > DROP VIEW v; > + > +--echo # > +--echo # MDEV-28152 Features for sequence > +--echo # > + > +create or replace sequence t1 as tinyint unsigned minvalue 1 maxvalue 2; > +show create sequence t1; > +show create table t1; > +select * from t1; > +select next value for t1; > +select next value for t1; > +--error ER_SEQUENCE_RUN_OUT > +select next value for t1; > + > +create or replace sequence t1 as tinyint unsigned minvalue 1 maxvalue 2 cycle; > +show create sequence t1; > +show create table t1; > +select * from t1; > +select next value for t1; > +select next value for t1; > +select next value for t1; > + > +--disable_ps_protocol > +create or replace sequence t1 minvalue -23 maxvalue 99999 as tinyint; > +--enable_ps_protocol > +show create sequence t1; > +show create table t1; > +select * from t1; > +select next value for t1; > +select next value for t1; > +select next value for t1; > + > +drop sequence t1; It is also interesting to see rolling over maximum value. > diff --git a/mysql-test/suite/sql_sequence/replication.result b/mysql-test/suite/sql_sequence/replication.result > index 94b1c72b9e0..2cfc246f2bf 100644 > --- a/mysql-test/suite/sql_sequence/replication.result > +++ b/mysql-test/suite/sql_sequence/replication.result > @@ -149,7 +149,7 @@ CREATE TABLE `s2` ( > ) ENGINE=myisam DEFAULT CHARSET=latin1 sequence=1; > show create sequence s2; > Table Create Table > -s2 CREATE SEQUENCE `s2` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +s2 CREATE SEQUENCE `s2` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > drop sequence s2; > ########################################### > select sequence syntax test > @@ -188,7 +188,7 @@ alter table s2 rename to s2_1; > rename table s2_1 to s2_2; > show create sequence s2_2; > Table Create Table > -s2_2 CREATE SEQUENCE `s2_2` start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > +s2_2 CREATE SEQUENCE `s2_2` as bigint start with 1 minvalue 1 maxvalue 9223372036854775806 increment by 1 cache 1000 nocycle ENGINE=MyISAM > select * from s2_2; > next_not_cached_value minimum_value maximum_value start_value increment cache_size cycle_option cycle_count > 1 1 9223372036854775806 1 1 1000 0 0 > diff --git a/mysql-test/suite/sql_sequence/slave_nextval.result b/mysql-test/suite/sql_sequence/slave_nextval.result > index bfbc472e117..36efa85d920 100644 > --- a/mysql-test/suite/sql_sequence/slave_nextval.result > +++ b/mysql-test/suite/sql_sequence/slave_nextval.result > @@ -4,7 +4,7 @@ CREATE SEQUENCE s; > INSERT INTO s VALUES (1,1,4,1,1,1,0,0); > show create sequence s; > Table Create Table > -s CREATE SEQUENCE `s` start with 1 minvalue 1 maxvalue 4 increment by 1 cache 1 nocycle ENGINE=MyISAM > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 4 increment by 1 cache 1 nocycle ENGINE=MyISAM > SELECT NEXTVAL(s); > NEXTVAL(s) > 1 > @@ -38,7 +38,7 @@ CREATE SEQUENCE s; > INSERT INTO s VALUES (1,1,3,1,1,1,1,0); > show create sequence s; > Table Create Table > -s CREATE SEQUENCE `s` start with 1 minvalue 1 maxvalue 3 increment by 1 cache 1 cycle ENGINE=MyISAM > +s CREATE SEQUENCE `s` as bigint start with 1 minvalue 1 maxvalue 3 increment by 1 cache 1 cycle ENGINE=MyISAM > SELECT NEXTVAL(s); > NEXTVAL(s) > 1 > diff --git a/sql/handler.h b/sql/handler.h > index 6b05da2ca98..f573dc835e9 100644 > --- a/sql/handler.h > +++ b/sql/handler.h > @@ -1055,6 +1055,7 @@ enum enum_schema_tables > SCH_PROCEDURES, > SCH_SCHEMATA, > SCH_SCHEMA_PRIVILEGES, > + SCH_SEQUENCES, > SCH_SESSION_STATUS, > SCH_SESSION_VARIABLES, > SCH_STATISTICS, > diff --git a/sql/sql_sequence.cc b/sql/sql_sequence.cc > index 11be2e1a691..c5754dc9670 100644 > --- a/sql/sql_sequence.cc > +++ b/sql/sql_sequence.cc > @@ -30,15 +30,6 @@ > #include "wsrep_mysqld.h" > #endif > > -struct Field_definition > -{ > - const char *field_name; > - uint length; > - const Type_handler *type_handler; > - LEX_CSTRING comment; > - ulong flags; > -}; > - > /* > Structure for all SEQUENCE tables > > @@ -48,30 +39,87 @@ struct Field_definition > a column named NEXTVAL. > */ > > +#define MAX_AUTO_INCREMENT_VALUE 65535 > + > +Sequence_row_definition sequence_structure(const Type_handler* handler) > +{ > + // We don't really care about src because it is unused in max_display_length_for_field(). > + const Conv_source src(handler, 0, system_charset_info); > + const uint32 len= handler->max_display_length_for_field(src) + 1; > + const LEX_CSTRING empty= {STRING_WITH_LEN("")}; > + const uint flag_unsigned= handler->is_unsigned() ? UNSIGNED_FLAG : 0; > #define FL (NOT_NULL_FLAG | NO_DEFAULT_VALUE_FLAG) > +#define FLV (NOT_NULL_FLAG | NO_DEFAULT_VALUE_FLAG | flag_unsigned) FL was a bad name (probably meant flag) and so you added even worse FLV (what it mean?) Please make both probably not long, but at least a bit meaningful. > + return {{{"next_not_cached_value", len, handler, empty, FLV}, > + {"minimum_value", len, handler, empty, FLV}, > + {"maximum_value", len, handler, empty, FLV}, > + {"start_value", len, handler, > + {STRING_WITH_LEN("start value when sequences is created or value " > + "if RESTART is used")}, FLV}, > + {"increment", 21, &type_handler_slonglong, > + {STRING_WITH_LEN("increment value")}, FL}, > + {"cache_size", 21, &type_handler_ulonglong, empty, > + FL | UNSIGNED_FLAG}, > + {"cycle_option", 1, &type_handler_utiny, > + {STRING_WITH_LEN("0 if no cycles are allowed, 1 if the sequence " "should begin a new cycle when maximum_value is " "passed")}, FL | UNSIGNED_FLAG}, > + {"cycle_count", 21, &type_handler_slonglong, > + {STRING_WITH_LEN("How many cycles have been done")}, FL}, > + {NULL, 0, &type_handler_slonglong, {STRING_WITH_LEN("")}, 0}}}; > +#undef FLV > +#undef FL > +} > + > +bool sequence_definition::is_allowed_value_type(enum_field_types type) > +{ > + switch (type) > + { > + case MYSQL_TYPE_TINY: > + case MYSQL_TYPE_SHORT: > + case MYSQL_TYPE_LONG: > + case MYSQL_TYPE_INT24: > + case MYSQL_TYPE_LONGLONG: > + return true; > + default: > + return false; > + } > +} > > -static Field_definition sequence_structure[]= > +Type_handler const *sequence_definition::value_type_handler() > { > - {"next_not_cached_value", 21, &type_handler_slonglong, > - {STRING_WITH_LEN("")}, FL}, > - {"minimum_value", 21, &type_handler_slonglong, {STRING_WITH_LEN("")}, FL}, > - {"maximum_value", 21, &type_handler_slonglong, {STRING_WITH_LEN("")}, FL}, > - {"start_value", 21, &type_handler_slonglong, {STRING_WITH_LEN("start value when sequences is created or value if RESTART is used")}, FL}, > - {"increment", 21, &type_handler_slonglong, > - {STRING_WITH_LEN("increment value")}, FL}, > - {"cache_size", 21, &type_handler_ulonglong, {STRING_WITH_LEN("")}, > - FL | UNSIGNED_FLAG}, > - {"cycle_option", 1, &type_handler_utiny, {STRING_WITH_LEN("0 if no cycles are allowed, 1 if the sequence should begin a new cycle when maximum_value is passed")}, > - FL | UNSIGNED_FLAG }, > - {"cycle_count", 21, &type_handler_slonglong, > - {STRING_WITH_LEN("How many cycles have been done")}, FL}, > - {NULL, 0, &type_handler_slonglong, {STRING_WITH_LEN("")}, 0} > -}; > + const Type_handler *handler= Type_handler::get_handler_by_field_type(value_type); > + return is_unsigned ? handler->type_handler_unsigned() : handler; > +} > > -#undef FL > +longlong sequence_definition::value_type_max() > +{ > + // value_type != MYSQL_TYPE_LONGLONG to avoid undefined behaviour > + // https://stackoverflow.com/questions/9429156/by-left-shifting-can-a-number-b… We use // only for one line coments please use /* */ above (please fx it in all your cases) > + return is_unsigned && value_type != MYSQL_TYPE_LONGLONG ? > + ~(~0ULL << 8 * value_type_handler()->calc_pack_length(0)) : > + ~value_type_min(); > +} > > +longlong sequence_definition::value_type_min() { > + return is_unsigned ? 0 : > + ~0ULL << (8 * value_type_handler()->calc_pack_length(0) - 1); > +} > > -#define MAX_AUTO_INCREMENT_VALUE 65535 > +/* > + Truncate `original` to `result`. > + If `original` is greater than value_type_max(), truncate down to value_type_max() > + If `original` is less than value_type_min(), truncate up to value_type_min() > +*/ Please use your standard description of all functions/methods /** One-line description Here longer multiline description of the function and maybe how it works @param paramater1 parameter1 description @param param2 param2 description @note Some notes about function and or implementation can be multiline. @return description what it return */ it also can be instead of @return returning value description @retval TRUE error @retval FALSE everything OK > +longlong sequence_definition::truncate_value(const Longlong_hybrid& original) > +{ > + if (is_unsigned) > + return original.to_ulonglong(value_type_max()); > + else if (original.is_unsigned_outside_of_signed_range()) > + return value_type_max(); > + else > + return original.value() > value_type_max() ? value_type_max() > + : original.value() < value_type_min() ? value_type_min() > + : original.value(); > +} > > /* > Check whether sequence values are valid. > @@ -82,49 +130,66 @@ static Field_definition sequence_structure[]= > true invalid > */ > > +// from_parser: whether to check foo_from_parser or foo, where foo in > +// {min_value, max_value, ...} Above method comment is in old style. You want to fix it (add a note), do just rewrite it please in the new style /** (see my example in this review) > bool sequence_definition::check_and_adjust(THD *thd, bool set_reserved_until) > { > longlong max_increment; > - DBUG_ENTER("sequence_definition::check"); > + DBUG_ENTER("sequence_definition::check_and_adjust"); > > if (!(real_increment= increment)) > real_increment= global_system_variables.auto_increment_increment; > > /* > - If min_value is not set, set it to LONGLONG_MIN or 1, depending on > + If min_value is not set, set it to value_type_min()+1 or 1, depending on > real_increment > */ > - if (!(used_fields & seq_field_used_min_value)) > - min_value= real_increment < 0 ? LONGLONG_MIN+1 : 1; > + if (!(used_fields & seq_field_specified_min_value)) > + min_value= real_increment < 0 ? value_type_min()+1 : 1; > + else > + { > + min_value= truncate_value(min_value_from_parser); > + if ((is_unsigned && (ulonglong) min_value <= (ulonglong) value_type_min()) || above line is too long (at least should fit in 80 better in 72 character) > + (!is_unsigned && min_value <= value_type_min())) > + { > + push_warning_printf( > + thd, Sql_condition::WARN_LEVEL_NOTE, ER_TRUNCATED_WRONG_VALUE, > + ER_THD(thd, ER_TRUNCATED_WRONG_VALUE), "INTEGER", "MINVALUE"); > + min_value= value_type_min() + 1; > + } > + } > > /* > - If max_value is not set, set it to LONGLONG_MAX or -1, depending on > + If max_value is not set, set it to value_type_max()-1 or -1, depending on > real_increment > */ > - if (!(used_fields & seq_field_used_max_value)) > - max_value= real_increment < 0 ? -1 : LONGLONG_MAX-1; > - > - if (max_value == LONGLONG_MAX) > - { > - push_warning_printf(thd, Sql_condition::WARN_LEVEL_NOTE, > - ER_TRUNCATED_WRONG_VALUE, > - ER_THD(thd, ER_TRUNCATED_WRONG_VALUE), > - "INTEGER", "MAXVALUE"); > - max_value= LONGLONG_MAX - 1; > - } > - if (min_value == LONGLONG_MIN) > + if (!(used_fields & seq_field_specified_max_value)) > + max_value= real_increment < 0 ? -1 : value_type_max()-1; > + else > { > - push_warning_printf( > - thd, Sql_condition::WARN_LEVEL_NOTE, ER_TRUNCATED_WRONG_VALUE, > - ER_THD(thd, ER_TRUNCATED_WRONG_VALUE), "INTEGER", "MINVALUE"); > - min_value= LONGLONG_MIN + 1; > + max_value= truncate_value(max_value_from_parser); > + if ((is_unsigned && (ulonglong) max_value >= (ulonglong) value_type_max()) || > + (!is_unsigned && max_value >= value_type_max())) > + { > + push_warning_printf( > + thd, Sql_condition::WARN_LEVEL_NOTE, ER_TRUNCATED_WRONG_VALUE, > + ER_THD(thd, ER_TRUNCATED_WRONG_VALUE), "INTEGER", "MAXVALUE"); > + max_value= value_type_max() - 1; > + } > } > > if (!(used_fields & seq_field_used_start)) > { > /* Use min_value or max_value for start depending on real_increment */ > start= real_increment < 0 ? max_value : min_value; > - } > + } else > + // If the supplied start value is out of range for the value type, > + // instead of immediately reporting error, we truncate it to > + // value_type_min or value_type_max depending on which side it is > + // one. Whenever such truncation happens, the condition that > + // max_value >= start >= min_value will be violated, and the error > + // will be reported then. We use // only for one line coments please use /* */ above (please fx it in all your cases) > + start= truncate_value(start_from_parser); Do we check that start value fit in [minvalue, maxvalue] range? > > if (set_reserved_until) > reserved_until= start; > @@ -136,12 +201,23 @@ bool sequence_definition::check_and_adjust(THD *thd, bool set_reserved_until) > llabs(real_increment) : > MAX_AUTO_INCREMENT_VALUE); > > - if (max_value >= start && > - max_value > min_value && > + // Common case for error, signed or unsigned. > + if (!is_allowed_value_type(value_type) || cache < 0) > + DBUG_RETURN(TRUE); > + > + // TODO: check for cache < (ULONGLONG_MAX - max_increment) / max_increment Why above is in TODO, IMHO it is not so difficult to make (maybe I do not see something?) > + if (is_unsigned && (ulonglong) max_value >= (ulonglong) start && > + (ulonglong) max_value > (ulonglong) min_value && > + (ulonglong) start >= (ulonglong) min_value && > + // Just like the case in signed, where a positive sequence > + // cannot have a negatvie increment, an unsigned sequence is > + // positive, so the increment has to be positive Why we can not count backward for unsigned sequence? Taking into account that increment probably should not exceed half of allowed range we always can store it as bigint (or decimal if we should support sequences with 1 value in them). We use // only for one line coments please use /* */ above (please fx it in all your cases) > + (real_increment > 0 && (ulonglong) reserved_until >= (ulonglong) min_value)) > + DBUG_RETURN(FALSE); > + > + if (!is_unsigned && max_value >= start && max_value > min_value && > start >= min_value && > - max_value != LONGLONG_MAX && > - min_value != LONGLONG_MIN && > - cache >= 0 && cache < (LONGLONG_MAX - max_increment) / max_increment && > + cache < (LONGLONG_MAX - max_increment) / max_increment && > ((real_increment > 0 && reserved_until >= min_value) || > (real_increment < 0 && reserved_until <= max_value))) > DBUG_RETURN(FALSE); > @@ -165,6 +241,11 @@ void sequence_definition::read_fields(TABLE *table) > cache= table->field[5]->val_int(); > cycle= table->field[6]->val_int(); > round= table->field[7]->val_int(); > + value_type= table->field[0]->type(); > + is_unsigned= table->field[0]->is_unsigned(); > + min_value_from_parser= Longlong_hybrid(min_value, is_unsigned); > + max_value_from_parser= Longlong_hybrid(max_value, is_unsigned); > + start_from_parser= Longlong_hybrid(start, is_unsigned); > dbug_tmp_restore_column_map(&table->read_set, old_map); > used_fields= ~(uint) 0; > print_dbug(); > @@ -181,10 +262,10 @@ void sequence_definition::store_fields(TABLE *table) > > /* zero possible delete markers & null bits */ > memcpy(table->record[0], table->s->default_values, table->s->null_bytes); > - table->field[0]->store(reserved_until, 0); > - table->field[1]->store(min_value, 0); > - table->field[2]->store(max_value, 0); > - table->field[3]->store(start, 0); > + table->field[0]->store(reserved_until, is_unsigned); > + table->field[1]->store(min_value, is_unsigned); > + table->field[2]->store(max_value, is_unsigned); > + table->field[3]->store(start, is_unsigned); > table->field[4]->store(increment, 0); > table->field[5]->store(cache, 0); > table->field[6]->store((longlong) cycle != 0, 0); > @@ -210,10 +291,17 @@ bool check_sequence_fields(LEX *lex, List<Create_field> *fields) > uint field_count; > uint field_no; > const char *reason; > + Sequence_row_definition row_structure; > DBUG_ENTER("check_sequence_fields"); > > field_count= fields->elements; > - if (field_count != array_elements(sequence_structure)-1) > + if (!field_count) > + { > + reason= "Wrong number of columns"; > + goto err; > + } > + row_structure= sequence_structure(fields->head()->type_handler()); > + if (field_count != array_elements(row_structure.fields)-1) > { > reason= "Wrong number of columns"; We do not have only english speaking users, it was done bad before and now we have chance to fix it and make it correct via localised errors. > goto err; > @@ -236,7 +324,7 @@ bool check_sequence_fields(LEX *lex, List<Create_field> *fields) > > for (field_no= 0; (field= it++); field_no++) > { > - Field_definition *field_def= &sequence_structure[field_no]; > + const Sequence_field_definition *field_def= &row_structure.fields[field_no]; > if (my_strcasecmp(system_charset_info, field_def->field_name, > field->field_name.str) || > field->flags != field_def->flags || > @@ -265,12 +353,13 @@ bool check_sequence_fields(LEX *lex, List<Create_field> *fields) > true Failure (out of memory) > */ > > -bool prepare_sequence_fields(THD *thd, List<Create_field> *fields) > +bool sequence_definition::prepare_sequence_fields(List<Create_field> *fields, bool alter) too long line > { > - Field_definition *field_info; > DBUG_ENTER("prepare_sequence_fields"); > + const Sequence_row_definition row_def= sequence_structure(value_type_handler()); We put DBUG_ENTER after variables definition (C code heritage) > > - for (field_info= sequence_structure; field_info->field_name ; field_info++) > + for (const Sequence_field_definition *field_info= row_def.fields; > + field_info->field_name; field_info++) > { > Create_field *new_field; > LEX_CSTRING field_name= {field_info->field_name, > @@ -285,6 +374,8 @@ bool prepare_sequence_fields(THD *thd, List<Create_field> *fields) > new_field->char_length= field_info->length; > new_field->comment= field_info->comment; > new_field->flags= field_info->flags; > + if (alter) > + new_field->change = field_name; > if (unlikely(fields->push_back(new_field))) > DBUG_RETURN(TRUE); /* purify inspected */ > } > @@ -313,20 +404,18 @@ bool sequence_insert(THD *thd, LEX *lex, TABLE_LIST *org_table_list) > Reprepare_observer *save_reprepare_observer; > sequence_definition *seq= lex->create_info.seq_create_info; > bool temporary_table= org_table_list->table != 0; > + /* > + seq is 0 if sequence was created with CREATE TABLE instead of > + CREATE SEQUENCE > + */ > + bool create_new= !seq; > Open_tables_backup open_tables_backup; > Query_tables_list query_tables_list_backup; > TABLE_LIST table_list; // For sequence table > DBUG_ENTER("sequence_insert"); > > - /* > - seq is 0 if sequence was created with CREATE TABLE instead of > - CREATE SEQUENCE > - */ > - if (!seq) > - { > - if (!(seq= new (thd->mem_root) sequence_definition)) > - DBUG_RETURN(TRUE); > - } > + if (create_new && !(seq= new (thd->mem_root) sequence_definition)) > + DBUG_RETURN(TRUE); > > #ifdef WITH_WSREP > if (WSREP_ON && seq->cache != 0) > @@ -387,7 +476,15 @@ bool sequence_insert(THD *thd, LEX *lex, TABLE_LIST *org_table_list) > else > table= org_table_list->table; > > - seq->reserved_until= seq->start; > + if (create_new) > + { > + seq->value_type= (*table->s->field)->type(); > + seq->is_unsigned= (*table->s->field)->is_unsigned(); > + // fixme: why do we need true here? because it is other copy so we reset it to the start, right? > + if (seq->check_and_adjust(thd, true)) > + DBUG_RETURN(TRUE); > + } > + > error= seq->write_initial_sequence(table); > { > uint save_unsafe_rollback_flags= > @@ -432,9 +529,9 @@ SEQUENCE::~SEQUENCE() > A sequence table can have many readers (trough normal SELECT's). > > We mark that we have a write lock in the table object so that > - ha_sequence::ha_write() can check if we have a lock. If already locked, then > + ha_sequence::write_row() can check if we have a lock. If already locked, then > ha_write() knows that we are running a sequence operation. If not, then > - ha_write() knows that it's an INSERT. > + ha_write() knows that it's an INSERT statement. > */ > > void SEQUENCE::write_lock(TABLE *table) > @@ -734,10 +831,9 @@ longlong SEQUENCE::next_value(TABLE *table, bool second_round, int *error) > write_lock(table); > > res_value= next_free_value; > - next_free_value= increment_value(next_free_value); > + next_free_value= increment_value(next_free_value, real_increment); > > - if ((real_increment > 0 && res_value < reserved_until) || > - (real_increment < 0 && res_value > reserved_until)) > + if (within_bounds(res_value, reserved_until, reserved_until, real_increment > 0)) > { > write_unlock(table); > DBUG_RETURN(res_value); > @@ -754,30 +850,10 @@ longlong SEQUENCE::next_value(TABLE *table, bool second_round, int *error) > overflow > */ > add_to= cache ? real_increment * cache : real_increment; > - out_of_values= 0; > > - if (real_increment > 0) > - { > - if (reserved_until > max_value - add_to || > - reserved_until + add_to > max_value) > - { > - reserved_until= max_value + 1; > - out_of_values= res_value >= reserved_until; > - } > - else > - reserved_until+= add_to; > - } > - else > - { > - if (reserved_until + add_to < min_value || > - reserved_until < min_value - add_to) > - { > - reserved_until= min_value - 1; > - out_of_values= res_value <= reserved_until; > - } > - else > - reserved_until+= add_to; > - } > + // TODO: consider extracting this refactoring to a separate earlier commit. You have send it on review, so I suppose you take decision about above? > + reserved_until= increment_value(reserved_until, add_to); > + out_of_values= !within_bounds(res_value, max_value + 1, min_value - 1, add_to > 0); > if (out_of_values) > { > if (!cycle || second_round) > @@ -866,7 +942,7 @@ int SEQUENCE::set_value(TABLE *table, longlong next_val, ulonglong next_round, > > write_lock(table); > if (is_used) > - next_val= increment_value(next_val); > + next_val= increment_value(next_val, real_increment); > > if (round > next_round) > goto end; // error = -1 > @@ -953,6 +1029,35 @@ bool Sql_cmd_alter_sequence::execute(THD *thd) > first_table)) > DBUG_RETURN(TRUE); > #endif /* WITH_WSREP */ > + > + if (new_seq->used_fields & seq_field_used_as) > + { > + // This shouldn't happen as it should have been prevented during > + // parsing. If something should not happened you better put DEBUG_ASSERT to notice in debug version if something goes wrong. Chenge comment stile if you want to leave it > + if (new_seq->used_fields - seq_field_used_as) > + DBUG_RETURN(TRUE); > + > + first_table->lock_type= TL_READ_NO_INSERT; > + first_table->mdl_request.set_type(MDL_SHARED_NO_WRITE); > + Alter_info alter_info; > + alter_info.flags= ALTER_CHANGE_COLUMN; > + if (new_seq->prepare_sequence_fields(&alter_info.create_list, true)) > + DBUG_RETURN(TRUE); > + Table_specification_st create_info; > + create_info.init(); > + create_info.alter_info= &alter_info; > + if (if_exists()) > + thd->push_internal_handler(&no_such_table_handler); > + error= mysql_alter_table(thd, &null_clex_str, &null_clex_str, &create_info, first_table, &alter_info, 0, (ORDER *) 0, 0, 0); > + if (if_exists()) > + { > + trapped_errors= no_such_table_handler.safely_trapped_errors(); > + thd->pop_internal_handler(); > + } > + // Do we need to store the sequence value in table share, like below? I do not understand comment above. (what it is about?) > + DBUG_RETURN(error); > + } > + > if (if_exists()) > thd->push_internal_handler(&no_such_table_handler); > error= open_and_lock_tables(thd, first_table, FALSE, 0); > @@ -989,22 +1094,30 @@ bool Sql_cmd_alter_sequence::execute(THD *thd) > if (!(new_seq->used_fields & seq_field_used_increment)) > new_seq->increment= seq->increment; > if (!(new_seq->used_fields & seq_field_used_min_value)) > - new_seq->min_value= seq->min_value; > + new_seq->min_value_from_parser= seq->min_value_from_parser; > if (!(new_seq->used_fields & seq_field_used_max_value)) > - new_seq->max_value= seq->max_value; > + new_seq->max_value_from_parser= seq->max_value_from_parser; > if (!(new_seq->used_fields & seq_field_used_start)) > - new_seq->start= seq->start; > + new_seq->start_from_parser= seq->start_from_parser; > if (!(new_seq->used_fields & seq_field_used_cache)) > new_seq->cache= seq->cache; > if (!(new_seq->used_fields & seq_field_used_cycle)) > new_seq->cycle= seq->cycle; > + if (!(new_seq->used_fields & seq_field_used_as)) > + { > + new_seq->value_type= seq->value_type; > + new_seq->is_unsigned= seq->is_unsigned; > + } > > /* If we should restart from a new value */ > if (new_seq->used_fields & seq_field_used_restart) > { > if (!(new_seq->used_fields & seq_field_used_restart_value)) > - new_seq->restart= new_seq->start; > - new_seq->reserved_until= new_seq->restart; > + new_seq->restart_from_parser= new_seq->start_from_parser; > + // Similar to start, we just need to truncate reserved_until and > + // the errors will be reported in check_and_adjust if truncation > + // happens on the wrong end. We use // only for one line coments please use /* */ above (please fx it in all your cases) > + new_seq->reserved_until= new_seq->truncate_value(new_seq->restart_from_parser); > } > > /* Let check_and_adjust think all fields are used */ > diff --git a/sql/sql_sequence.h b/sql/sql_sequence.h > index e2968cc20ae..0de92a90639 100644 > --- a/sql/sql_sequence.h > +++ b/sql/sql_sequence.h > @@ -25,12 +25,35 @@ > #define seq_field_used_cycle 32 > #define seq_field_used_restart 64 > #define seq_field_used_restart_value 128 > +#define seq_field_used_as 256 > +#define seq_field_specified_min_value 512 > +#define seq_field_specified_max_value 1024 > > /* Field position in sequence table for some fields we refer to directly */ > #define NEXT_FIELD_NO 0 > #define MIN_VALUE_FIELD_NO 1 > #define ROUND_FIELD_NO 7 > > +#include "mysql_com.h" > +#include "sql_type_int.h" > + > +class Create_field; > +class Type_handler; > + > +struct Sequence_field_definition > +{ > + const char *field_name; > + uint length; > + const Type_handler *type_handler; > + LEX_CSTRING comment; > + ulong flags; > +}; > + > +struct Sequence_row_definition > +{ > + Sequence_field_definition fields[9]; > +}; > + > /** > sequence_definition is used when defining a sequence as part of create > */ > @@ -39,20 +62,37 @@ class sequence_definition :public Sql_alloc > { > public: > sequence_definition(): > - min_value(1), max_value(LONGLONG_MAX-1), start(1), increment(1), > - cache(1000), round(0), restart(0), cycle(0), used_fields(0) > + min_value_from_parser(1, false), > + max_value_from_parser(LONGLONG_MAX-1, false), start_from_parser(1, false), > + increment(1), cache(1000), round(0), restart_from_parser(0, false), cycle(0), used_fields(0), > + // We use value type and is_unsigned instead of a handler because > + // Type_handler is incomplete, which we cannot initialise here > + // with &type_handler_slonglong. We use // only for one line coments please use /* */ above (please fx it in all your cases) > + value_type(MYSQL_TYPE_LONGLONG), is_unsigned(false) > {} > longlong reserved_until; > longlong min_value; > longlong max_value; > longlong start; > + Longlong_hybrid min_value_from_parser; > + Longlong_hybrid max_value_from_parser; > + Longlong_hybrid start_from_parser; > longlong increment; > longlong cache; > ulonglong round; > + // TODO: allow unsigned in restart > longlong restart; // alter sequence restart value > + Longlong_hybrid restart_from_parser; > bool cycle; > uint used_fields; // Which fields where used in CREATE > - > + enum_field_types value_type; // value type of the sequence > + bool is_unsigned; > + > + Type_handler const *value_type_handler(); > + // max value for the value type, e.g. 32767 for smallint. > + longlong value_type_max(); > + // min value for the value type, e.g. -32768 for smallint. > + longlong value_type_min(); > bool check_and_adjust(THD *thd, bool set_reserved_until); > void store_fields(TABLE *table); > void read_fields(TABLE *table); > @@ -60,12 +100,16 @@ class sequence_definition :public Sql_alloc > int write(TABLE *table, bool all_fields); > /* This must be called after sequence data has been updated */ > void adjust_values(longlong next_value); > + longlong truncate_value(const Longlong_hybrid& original); > inline void print_dbug() > { > DBUG_PRINT("sequence", ("reserved: %lld start: %lld increment: %lld min_value: %lld max_value: %lld cache: %lld round: %lld", > reserved_until, start, increment, min_value, > max_value, cache, round)); > } > + static bool is_allowed_value_type(enum_field_types type); > + bool prepare_sequence_fields(List<Create_field> *fields, bool alter); > + > protected: > /* > The following values are the values from sequence_definition > @@ -107,24 +151,54 @@ class SEQUENCE :public sequence_definition > longlong next_value(TABLE *table, bool second_round, int *error); > int set_value(TABLE *table, longlong next_value, ulonglong round_arg, > bool is_used); > - longlong increment_value(longlong value) > + bool within_bounds(const longlong value, const longlong upper, const longlong lower, bool increasing) > + { > + return > + (is_unsigned && increasing && (ulonglong) value < (ulonglong) upper) || > + (is_unsigned && !increasing && (ulonglong) value > (ulonglong) lower) || > + (!is_unsigned && increasing && value < upper) || > + (!is_unsigned && !increasing && value > lower); > + } > + > + longlong increment_value(longlong value, const longlong increment) > { > - if (real_increment > 0) > + if (is_unsigned) > { > - if (value > max_value - real_increment || > - value + real_increment > max_value) > - value= max_value + 1; > + if (increment > 0) > + { > + // in case value + increment overflows > + if ((ulonglong) value > (ulonglong) max_value - (ulonglong) increment || > + // in case max_value - increment underflows > + (ulonglong) value + (ulonglong) increment > (ulonglong) max_value) > + value= max_value + 1; > + else > + value+= increment; > + } > else > - value+= real_increment; > - } > - else > - { > - if (value + real_increment < min_value || > - value < min_value - real_increment) > - value= min_value - 1; > + { > + if ((ulonglong) value - (ulonglong) (-increment) < (ulonglong) min_value || > + (ulonglong) value < (ulonglong) min_value + (ulonglong) (- increment)) > + value= min_value - 1; > + else > + value+= increment; > + } > + } else > + if (increment > 0) > + { > + if (value > max_value - increment || > + value + increment > max_value) > + value= max_value + 1; > + else > + value+= increment; > + } > else > - value+= real_increment; > - } > + { > + if (value + increment < min_value || > + value < min_value - increment) > + value= min_value - 1; > + else > + value+= increment; > + } > return value; > } > > @@ -159,9 +233,6 @@ class SEQUENCE_LAST_VALUE > uchar table_version[MY_UUID_SIZE]; > }; > > - > -class Create_field; > -extern bool prepare_sequence_fields(THD *thd, List<Create_field> *fields); > extern bool check_sequence_fields(LEX *lex, List<Create_field> *fields); > extern bool sequence_insert(THD *thd, LEX *lex, TABLE_LIST *table_list); > #endif /* SQL_SEQUENCE_INCLUDED */ > diff --git a/sql/sql_show.cc b/sql/sql_show.cc > index 1fd31bd4947..12e4f655515 100644 > --- a/sql/sql_show.cc > +++ b/sql/sql_show.cc > @@ -2694,12 +2694,29 @@ static int show_create_sequence(THD *thd, TABLE_LIST *table_list, > > packet->append(STRING_WITH_LEN("CREATE SEQUENCE ")); > append_identifier(thd, packet, &alias); > + /* Do not show " as <type>" in oracle mode as it is not supported: In case of multiline comment /* should be on separate line > + https://docs.oracle.com/en/database/oracle/oracle-database/21/sqlrf/CREATE-… you also probably should not show it if it is default (BIGINT), see other comment about it in the beginning. BTW add test of making sequence with/without oracle mode and dumping and restoring (mysqldump) them for all types. You can find examples of the tests by "ls mysql-test/main/*dump*.test". > + */ > + if (!(sql_mode & MODE_ORACLE)) > + { > + packet->append(STRING_WITH_LEN(" as ")); > + packet->append(seq->value_type_handler()->name().lex_cstring()); > + } > packet->append(STRING_WITH_LEN(" start with ")); > - packet->append_longlong(seq->start); > + if (seq->is_unsigned) > + packet->append_ulonglong(seq->start); > + else > + packet->append_longlong(seq->start); > packet->append(STRING_WITH_LEN(" minvalue ")); > - packet->append_longlong(seq->min_value); > + if (seq->is_unsigned) > + packet->append_ulonglong(seq->min_value); > + else > + packet->append_longlong(seq->min_value); > packet->append(STRING_WITH_LEN(" maxvalue ")); > - packet->append_longlong(seq->max_value); > + if (seq->is_unsigned) > + packet->append_ulonglong(seq->max_value); > + else > + packet->append_longlong(seq->max_value); > packet->append(STRING_WITH_LEN(" increment by ")); > packet->append_longlong(seq->increment); > if (seq->cache) > @@ -5512,6 +5529,36 @@ int fill_schema_schemata(THD *thd, TABLE_LIST *tables, COND *cond) > DBUG_RETURN(0); > } > > +static int get_schema_sequence_record(THD *thd, TABLE_LIST *tables, > + TABLE *table, bool res, > + const LEX_CSTRING *db_name, > + const LEX_CSTRING *table_name) > +{ > + DBUG_ENTER("get_sequence_record"); > + CHARSET_INFO *cs= system_charset_info; > + restore_record(table, s->default_values); > + sequence_definition *seq= tables->table->s->sequence; > + if (tables->table->s->table_type == TABLE_TYPE_SEQUENCE) > + { > + const Type_handler *handler= seq->value_type_handler(); > + table->field[0]->store(STRING_WITH_LEN("def"), cs); > + table->field[1]->store(db_name->str, db_name->length, cs); > + table->field[2]->store(table_name->str, table_name->length, cs); > + table->field[3]->store(handler->name().lex_cstring(), cs); > + table->field[4]->store(8 * handler->calc_pack_length(0)); > + table->field[5]->store(2); > + table->field[5]->set_notnull(); > + table->field[6]->store(0); > + table->field[6]->set_notnull(); > + table->field[7]->store(seq->start, seq->is_unsigned); > + table->field[8]->store(seq->min_value, seq->is_unsigned); > + table->field[9]->store(seq->max_value, seq->is_unsigned); > + table->field[10]->store(seq->increment, 0); > + table->field[11]->store(seq->cycle); > + DBUG_RETURN(schema_table_store_record(thd, table)); > + } > + DBUG_RETURN(0); > +} > > static int get_schema_tables_record(THD *thd, TABLE_LIST *tables, > TABLE *table, bool res, > @@ -9419,6 +9466,29 @@ ST_FIELD_INFO proc_fields_info[]= > }; > > > +ST_FIELD_INFO sequence_fields_info[]= > +{ > + Column("SEQUENCE_CATALOG", Catalog(), NOT_NULL, OPEN_FRM_ONLY), > + Column("SEQUENCE_SCHEMA", Name(), NOT_NULL, OPEN_FRM_ONLY), > + Column("SEQUENCE_NAME", Name(), NOT_NULL, "Table", OPEN_FRM_ONLY), > + Column("DATA_TYPE", Name(), NOT_NULL), > + Column("NUMERIC_PRECISION", SLong(21), NOT_NULL), > + Column("NUMERIC_PRECISION_RADIX", SLong(21), NULLABLE), > + Column("NUMERIC_SCALE", SLong(21), NULLABLE), > + // Decimal types for these values to incorporate possibly unsigned > + // longlongs. We use // only for one line coments please use /* */ above (please fx it in all your cases) > + Column("START_VALUE", Decimal(2100), NOT_NULL), > + Column("MINIMUM_VALUE", Decimal(2100), NOT_NULL), > + Column("MAXIMUM_VALUE", Decimal(2100), NOT_NULL), > + Column("INCREMENT", SLonglong(21), NOT_NULL), > + Column("CYCLE_OPTION", SLonglong(21), NOT_NULL), > + Column("DECLARED_DATA_TYPE", SLong(21), NULLABLE), > + Column("DECLARED_NUMERIC_PRECISION", SLong(21), NULLABLE), > + Column("DECLARED_NUMERIC_SCALE", SLong(21), NULLABLE), > + CEnd() > +}; > + > + > ST_FIELD_INFO stat_fields_info[]= > { > Column("TABLE_CATALOG", Catalog(), NOT_NULL, OPEN_FRM_ONLY), > @@ -9985,6 +10055,8 @@ ST_SCHEMA_TABLE schema_tables[]= > fill_schema_schemata, make_schemata_old_format, 0, 1, -1, 0, 0}, > {"SCHEMA_PRIVILEGES", Show::schema_privileges_fields_info, 0, > fill_schema_schema_privileges, 0, 0, -1, -1, 0, 0}, > + {"SEQUENCES", Show::sequence_fields_info, 0, > + get_all_tables, make_old_format, get_schema_sequence_record, 1, 2, 0, 0}, > {"SESSION_STATUS", Show::variables_fields_info, 0, > fill_status, make_old_format, 0, 0, -1, 0, 0}, > {"SESSION_VARIABLES", Show::variables_fields_info, 0, > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy > index e3298a4a6c1..27ce8bcdbfc 100644 > --- a/sql/sql_yacc.yy > +++ b/sql/sql_yacc.yy > @@ -200,6 +200,9 @@ void _CONCAT_UNDERSCORED(turn_parser_debug_on,yyparse)() > ulonglong ulonglong_number; > longlong longlong_number; > uint sp_instr_addr; > + // Longlong_hybrid does not have a default constructor, hence the > + // default value below. We use // only for one line coments please use /* */ above (please fx it in all your cases) > + Longlong_hybrid longlong_hybrid_number= Longlong_hybrid(0, false); > > /* structs */ > LEX_CSTRING lex_str; > @@ -1466,7 +1469,10 @@ bool my_yyoverflow(short **a, YYSTYPE **b, size_t *yystacksize); > ulonglong_num real_ulonglong_num > > %type <longlong_number> > - sequence_value_num sequence_truncated_value_num > + sequence_value_num > + > +%type <longlong_hybrid_number> > + sequence_value_hybrid_num sequence_truncated_value_hybrid_num > > %type <choice> choice > > @@ -2429,8 +2435,8 @@ create: > } > > /* No fields specified, generate them */ > - if (unlikely(prepare_sequence_fields(thd, > - &lex->alter_info.create_list))) > + if (unlikely(lex->create_info.seq_create_info->prepare_sequence_fields( > + &lex->alter_info.create_list, false))) > MYSQL_YYABORT; > > /* CREATE SEQUENCE always creates a sequence */ > @@ -2605,13 +2611,25 @@ sequence_defs: > ; > > sequence_def: > - MINVALUE_SYM opt_equal sequence_truncated_value_num > + AS int_type field_options > + { > + if (unlikely(Lex->create_info.seq_create_info->used_fields & > + seq_field_used_as)) > + my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "AS")); > + if ($3 & ZEROFILL_FLAG) > + my_yyabort_error((ER_BAD_OPTION_VALUE, MYF(0), "ZEROFILL", "AS")); > + Lex->create_info.seq_create_info->value_type = $2->field_type(); > + Lex->create_info.seq_create_info->is_unsigned = $3 & UNSIGNED_FLAG ? true : false; > + Lex->create_info.seq_create_info->used_fields|= seq_field_used_as; > + } > + | MINVALUE_SYM opt_equal sequence_truncated_value_hybrid_num > { > if (unlikely(Lex->create_info.seq_create_info->used_fields & > seq_field_used_min_value)) > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "MINVALUE")); > - Lex->create_info.seq_create_info->min_value= $3; > + Lex->create_info.seq_create_info->min_value_from_parser= $3; > Lex->create_info.seq_create_info->used_fields|= seq_field_used_min_value; > + Lex->create_info.seq_create_info->used_fields|= seq_field_specified_min_value; > } > | NO_SYM MINVALUE_SYM > { > @@ -2625,13 +2643,14 @@ sequence_def: > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "MINVALUE")); > Lex->create_info.seq_create_info->used_fields|= seq_field_used_min_value; > } > - | MAXVALUE_SYM opt_equal sequence_truncated_value_num > + | MAXVALUE_SYM opt_equal sequence_truncated_value_hybrid_num > { > if (unlikely(Lex->create_info.seq_create_info->used_fields & > seq_field_used_max_value)) > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "MAXVALUE")); > - Lex->create_info.seq_create_info->max_value= $3; > + Lex->create_info.seq_create_info->max_value_from_parser= $3; > Lex->create_info.seq_create_info->used_fields|= seq_field_used_max_value; > + Lex->create_info.seq_create_info->used_fields|= seq_field_specified_max_value; > } > | NO_SYM MAXVALUE_SYM > { > @@ -2645,12 +2664,12 @@ sequence_def: > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "MAXVALUE")); > Lex->create_info.seq_create_info->used_fields|= seq_field_used_max_value; > } > - | START_SYM opt_with sequence_value_num > + | START_SYM opt_with sequence_value_hybrid_num > { > if (unlikely(Lex->create_info.seq_create_info->used_fields & > seq_field_used_start)) > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "START")); > - Lex->create_info.seq_create_info->start= $3; > + Lex->create_info.seq_create_info->start_from_parser= $3; > Lex->create_info.seq_create_info->used_fields|= seq_field_used_start; > } > | INCREMENT_SYM opt_by sequence_value_num > @@ -2705,7 +2724,7 @@ sequence_def: > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "RESTART")); > Lex->create_info.seq_create_info->used_fields|= seq_field_used_restart; > } > - | RESTART_SYM opt_with sequence_value_num > + | RESTART_SYM opt_with sequence_value_hybrid_num > { > if (unlikely(Lex->sql_command != SQLCOM_ALTER_SEQUENCE)) > { > @@ -2715,7 +2734,7 @@ sequence_def: > if (unlikely(Lex->create_info.seq_create_info->used_fields & > seq_field_used_restart)) > my_yyabort_error((ER_DUP_ARGUMENT, MYF(0), "RESTART")); > - Lex->create_info.seq_create_info->restart= $3; > + Lex->create_info.seq_create_info->restart_from_parser= $3; > Lex->create_info.seq_create_info->used_fields|= seq_field_used_restart | seq_field_used_restart_value; > } > ; > @@ -7122,6 +7141,8 @@ alter: > { > /* Create a generic ALTER SEQUENCE statment. */ > Lex->m_sql_cmd= new (thd->mem_root) Sql_cmd_alter_sequence($3); > + if ((Lex->create_info.seq_create_info->used_fields & seq_field_used_as) && (Lex->create_info.seq_create_info->used_fields - seq_field_used_as)) Above lie is to long > + my_yyabort_error((ER_NOT_SUPPORTED_YET, MYF(0), "ALTER SEQUENCE with both AS <type> and something else.")); Above lie is to long > if (unlikely(Lex->m_sql_cmd == NULL)) > MYSQL_YYABORT; > } stmt_end {} > @@ -12563,6 +12584,7 @@ real_ulong_num: > | dec_num_error { MYSQL_YYABORT; } > ; > > +// For simple sequence metadata values that are signed and do not need truncation Above lie is to long > sequence_value_num: > opt_plus NUM { int error; $$= (longlong) my_strtoll10($2.str, (char**) 0, &error); } > | opt_plus LONG_NUM { int error; $$= (longlong) my_strtoll10($2.str, (char**) 0, &error); } > @@ -12579,15 +12601,74 @@ sequence_value_num: > } > ; > > -sequence_truncated_value_num: > - opt_plus NUM { int error; $$= (longlong) my_strtoll10($2.str, (char**) 0, &error); } > - | opt_plus LONG_NUM { int error; $$= (longlong) my_strtoll10($2.str, (char**) 0, &error); } > - | opt_plus ULONGLONG_NUM { $$= LONGLONG_MAX; } > - | opt_plus DECIMAL_NUM { $$= LONGLONG_MAX; } > - | '-' NUM { int error; $$= -(longlong) my_strtoll10($2.str, (char**) 0, &error); } > - | '-' LONG_NUM { int error; $$= -(longlong) my_strtoll10($2.str, (char**) 0, &error); } > - | '-' ULONGLONG_NUM { $$= LONGLONG_MIN; } > - | '-' DECIMAL_NUM { $$= LONGLONG_MIN; } > +// For sequence metadata values that may be unsigned but do not need truncation (start, restart) Above lie is to long (it looks like you have them a lot, so please fix all lines to fit in 80 (better 72 character). > +sequence_value_hybrid_num: > + opt_plus NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | opt_plus LONG_NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | opt_plus ULONGLONG_NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), true); > + } > + | '-' NUM > + { > + int error; > + $$= Longlong_hybrid(- my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | '-' LONG_NUM > + { > + int error; > + $$= Longlong_hybrid(- my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | '-' ULONGLONG_NUM > + { > + int error; > + const ulonglong abs= my_strtoll10($2.str, (char**) 0, &error); > + if (abs == 1 + (ulonglong) LONGLONG_MAX) > + $$= Longlong_hybrid(LONGLONG_MIN, false); > + else > + thd->parse_error(ER_DATA_OUT_OF_RANGE); > + } > + ; > + > +// For sequence metadata values that may be unsigned and need truncation (maxvalue, minvalue) > +sequence_truncated_value_hybrid_num: > + opt_plus NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | opt_plus LONG_NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | opt_plus ULONGLONG_NUM > + { > + int error; > + $$= Longlong_hybrid(my_strtoll10($2.str, (char**) 0, &error), true); > + } > + | opt_plus DECIMAL_NUM { $$= Longlong_hybrid(ULONGLONG_MAX, true); } > + | '-' NUM > + { > + int error; > + $$= Longlong_hybrid(- my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | '-' LONG_NUM > + { > + int error; > + $$= Longlong_hybrid(- my_strtoll10($2.str, (char**) 0, &error), false); > + } > + | '-' ULONGLONG_NUM { $$= Longlong_hybrid(LONGLONG_MIN, false); } > + | '-' DECIMAL_NUM { $$= Longlong_hybrid(LONGLONG_MIN, false); } > ; > > ulonglong_num:

3 6

Re: [Maria-developers] 8c1ad2a9fe9: MDEV-30633 DATETIME to TIMESTAMP conversion to return maximum timestamp on overflow
by Sergei Golubchik 08 Mar '23

08 Mar '23

Hi, Alexander, please add tests with fractional seconds. remember, that in UTC the max timestamp(2) is 2038-01-19 03:14:07.99 and max timestamp(6) is 2038-01-19 03:14:07.999999 On Mar 08, Alexander Barkov wrote: > revision-id: 8c1ad2a9fe9 (mariadb-10.11.1-48-g8c1ad2a9fe9) > parent(s): ce4a289f1c3 > author: Alexander Barkov > committer: Alexander Barkov > timestamp: 2023-02-13 17:25:18 +0400 > message: > > MDEV-30633 DATETIME to TIMESTAMP conversion to return maximum timestamp on overflow a bit more verbose description would be nice > diff --git a/mysql-test/main/events_bugs.result b/mysql-test/main/events_bugs.result > --- a/mysql-test/main/events_bugs.result > +++ b/mysql-test/main/events_bugs.result > @@ -35,12 +35,34 @@ SET NAMES latin1; > set @a=3; > CREATE PROCEDURE p_16 () CREATE EVENT e_16 ON SCHEDULE EVERY @a SECOND DO SET @a=5; > ERROR HY000: Recursion of EVENT DDL statements is forbidden when body is present > +SET time_zone='+00:00'; > +SET timestamp=UNIX_TIMESTAMP('2023-02-13 00:00:00'); > create event e_55 on schedule at 99990101000000 do drop table t; > -ERROR HY000: Incorrect AT value: '99990101000000' Hmm, why did you not keep it an error? > +Warnings: > +Warning 1292 Truncated incorrect timestamp value: '9999-01-01 00:00:00' > +Warning 1105 Event scheduler is switched off, use SET GLOBAL event_scheduler=ON to enable it. > diff --git a/mysql-test/main/func_time.result b/mysql-test/main/func_time.result > --- a/mysql-test/main/func_time.result > +++ b/mysql-test/main/func_time.result > @@ -599,19 +599,25 @@ Warnings: > Warning 1292 Truncated incorrect unixtime value: '2147483648' > select unix_timestamp('2039-01-20 01:00:00'); > unix_timestamp('2039-01-20 01:00:00') > -NULL > +2147483647 > +Warnings: > +Warning 1292 Truncated incorrect timestamp value: '2039-01-20 01:00:00' it's consistent with MariaDB [test]> select cast(1e30 as int); +---------------------+ | cast(1e30 as int) | +---------------------+ | 9223372036854775807 | +---------------------+ 1 row in set, 1 warning (0.000 sec) Note (Code 1916): Got overflow when converting '1e30' to SIGNED BIGINT. Value truncated good. Except that it's Note vs Warning. Want to unify this somehow? > select unix_timestamp('1968-01-20 01:00:00'); > unix_timestamp('1968-01-20 01:00:00') > NULL isn't it strange? That overflow is capped to max value, while underflow is NULL? > diff --git a/mysql-test/main/select.result b/mysql-test/main/select.result > --- a/mysql-test/main/select.result > +++ b/mysql-test/main/select.result > @@ -3766,11 +3766,15 @@ AND t1.ts BETWEEN "2006-01-01" AND "2006-12-31"; > id select_type table type possible_keys key key_len ref rows Extra > 1 SIMPLE t2 const PRIMARY PRIMARY 4 const 1 > 1 SIMPLE t1 range ts ts 4 NULL 2 Using index condition; Using where > +Warnings: > +Warning 1292 Truncated incorrect timestamp value: '2999-12-31 00:00:00' Looks questionable. The query compares timestamp with a datetime. I think in this case timestamp should be casted to a datetime (so, no warning), not vice versa. > SELECT * FROM t1 LEFT JOIN t2 ON (t1.a=t2.a) WHERE t1.a=30 > AND t1.ts BETWEEN t2.dt1 AND t2.dt2 > AND t1.ts BETWEEN "2006-01-01" AND "2006-12-31"; > a ts a dt1 dt2 > 30 2006-01-03 23:00:00 30 2006-01-01 00:00:00 2999-12-31 00:00:00 > +Warnings: > +Warning 1292 Truncated incorrect timestamp value: '2999-12-31 00:00:00' > DROP TABLE t1,t2; > create table t1 (a bigint unsigned); > insert into t1 values > diff --git a/sql/item_timefunc.cc b/sql/item_timefunc.cc > --- a/sql/item_timefunc.cc > +++ b/sql/item_timefunc.cc > @@ -2793,11 +2793,17 @@ bool Item_func_convert_tz::get_date(THD *thd, MYSQL_TIME *ltime, > return true; > > { > - uint not_used; > - my_time_tmp= from_tz->TIME_to_gmt_sec(ltime, &not_used); > + uint error_code; > + my_time_tmp= from_tz->TIME_to_gmt_sec(ltime, &error_code); > ulong sec_part= ltime->second_part; > - /* my_time_tmp is guaranteed to be in the allowed range */ > - if (my_time_tmp) > + /* > + my_time_tmp is guaranteed to be in the allowed range. > + Don't perform the conversion in case the source DATETIME was above > + TIMESTAMP_MAX_VALUE (and was truncated to TIMESTAMP_MAX_VALUE). why not? > + */ > + if (my_time_tmp && > + (my_time_tmp != TIMESTAMP_MAX_VALUE || > + error_code != ER_WARN_DATA_OUT_OF_RANGE)) why `my_time_tmp != TIMESTAMP_MAX_VALUE` ? shouldn't error_code check be sufficient? > to_tz->gmt_sec_to_TIME(ltime, my_time_tmp); > /* we rely on the fact that no timezone conversion can change sec_part */ > ltime->second_part= sec_part; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 6e28a576913: move alloca() definition from all *.h files to one place
by Sergei Golubchik 10 Feb '23

10 Feb '23

Hi, Julius, this is the combined diff of both commits: git diff e1ce867e5ab0c^^ e1ce867e5ab0c On Feb 10, Julius Goryavsky wrote: > diff --git a/include/CMakeLists.txt b/include/CMakeLists.txt > index a7b98a11050..7d87c4a42b3 100644 > --- a/include/CMakeLists.txt > +++ b/include/CMakeLists.txt > @@ -104,6 +105,7 @@ ENDMACRO() > > INSTALL_COMPAT_HEADER(my_global.h "") > INSTALL_COMPAT_HEADER(my_config.h "") > +INSTALL_COMPAT_HEADER(my_alloca.h "") I wouldn't use INSTALL_COMPAT_HEADER for my_alloca.h > INSTALL_COMPAT_HEADER(my_sys.h "") > INSTALL_COMPAT_HEADER(mysql_version.h " > #include <mariadb_version.h> > diff --git a/include/my_alloca.h b/include/my_alloca.h > new file mode 100644 > index 00000000000..fec0320195e > --- /dev/null > +++ b/include/my_alloca.h > @@ -0,0 +1,37 @@ > +/* Copyright (c) 2023, MariaDB Corporation. > + > + This program is free software; you can redistribute it and/or modify > + it under the terms of the GNU General Public License as published by > + the Free Software Foundation; version 2 of the License. > + > + This program is distributed in the hope that it will be useful, > + but WITHOUT ANY WARRANTY; without even the implied warranty of > + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + GNU General Public License for more details. > + > + You should have received a copy of the GNU General Public License > + along with this program; if not, write to the Free Software > + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA */ > + > +#ifndef MY_ALLOCA_INCLUDED > +#define MY_ALLOCA_INCLUDED > + > +#ifdef _WIN32 > +#include <malloc.h> /*for alloca*/ > +#ifndef alloca When can alloca be defined? > +#define alloca _alloca > +#endif > +#else > +#include <stdlib.h> why stdlib.h? > +#ifdef HAVE_ALLOCA_H > +#include <alloca.h> > +#endif > +#endif > + > +#if defined(HAVE_ALLOCA) > +#if defined(__GNUC__) && !defined(HAVE_ALLOCA_H) && !defined(alloca) add a comment describing your MinGW use case when this #if is needed > +#define alloca __builtin_alloca > +#endif /* GNUC */ > +#endif > + > +#endif /* MY_ALLOCA_INCLUDED */ > diff --git a/include/mysql/service_encryption.h b/include/mysql/service_encryption.h > index 69d205a27e8..cab1418e1d6 100644 > --- a/include/mysql/service_encryption.h > +++ b/include/mysql/service_encryption.h > @@ -24,21 +24,19 @@ > *provider* (encryption plugin). > */ > > +#ifndef MYSQL_ABI_CHECK > +#include <my_alloca.h> > +#endif > + > #ifdef __cplusplus > extern "C" { > #endif > > #ifndef MYSQL_ABI_CHECK why not to put #include <my_alloca.h> here? > #ifdef _WIN32 > -#include <malloc.h> > #ifndef __cplusplus > #define inline __inline > #endif > -#else > -#include <stdlib.h> I suspect stdlib.h belongs here, not in my_alloca.h > -#ifdef HAVE_ALLOCA_H > -#include <alloca.h> > -#endif > #endif > #endif Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] ceb75e9bd9e: MDEV-30203: Move mysql symlinks to different package (fixes)
by Sergei Golubchik 10 Feb '23

10 Feb '23

Hi, Daniel, On Feb 09, Daniel Black wrote: > revision-id: ceb75e9bd9e (mariadb-10.11.1-152-gceb75e9bd9e) > parent(s): 526fce28b90 > author: Daniel Black > committer: Daniel Black > timestamp: 2023-02-09 12:20:57 +1100 > message: > > MDEV-30203: Move mysql symlinks to different package (fixes) > > Fixes MDEV-30571 MariaDB-client does not pull MariaDB-compat anymore, conflicts with RHEL packages > > Fixes MDEV-30574 MariaDB-server conflicts with MariaDB-common from previous versions > > MariaDB-Server now depends on MariaDB-common 11.0+. > > MariaDB-{Client,Server} append to the Requirements rather than overriding them. > > MariaDB-Server package now depends on the MariaDB-client 11.0+. > > Because of moves between packages MariaDB-Client conflicts with > MariaDB-server < 11.0.0. > > diff --git a/cmake/cpack_rpm.cmake b/cmake/cpack_rpm.cmake > index 638f6b2dabb..e88a5493eac 100644 > --- a/cmake/cpack_rpm.cmake > +++ b/cmake/cpack_rpm.cmake > @@ -189,7 +189,7 @@ FOREACH(SYM_COMPONENT Server Client) > SET(CPACK_RPM_${SYM}_PACKAGE_ARCHITECTURE "noarch") > SET(CPACK_RPM_${SYM}_USER_FILELIST ${ignored}) > STRING(TOLOWER ${SYM_COMPONENT} SYM_COMPONENT_LOWER) > - SET(CPACK_RPM_${SYM}_PACKAGE_REQUIRES "MariaDB-${SYM_COMPONENT_LOWER} >= 11.0.0") > + SETA(CPACK_RPM_${SYM}_PACKAGE_REQUIRES "MariaDB-${SYM_COMPONENT_LOWER} >= 11.0.0") why? > SETA(CPACK_RPM_${SYM_COMPONENT_LOWER}_PACKAGE_RECOMMENDS "MariaDB-${SYM}") > ENDFOREACH() > > @@ -205,10 +205,10 @@ SETA(CPACK_RPM_client_PACKAGE_PROVIDES > "mysql-client" > "mytop") > SETA(CPACK_RPM_client_PACKAGE_CONFLICTS > - "MariaDB-server < 10.6.0") > + "MariaDB-server < 11.0.0") because of resolve-stack-dump, ok > SETA(CPACK_RPM_common_PACKAGE_CONFLICTS > - "MariaDB-server < 10.6.1") > + "MariaDB-server < 11.0.0") why? > SETA(CPACK_RPM_devel_PACKAGE_OBSOLETES > "MySQL-devel") Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

[Maria-developers] MDEV-30218: On what should be passed as 'rows' to filtering code.
by Sergey Petrunia 01 Feb '23

01 Feb '23

Hello Igor, So I'm looking at the bb-10.4-release-igor tree with your latest patch: commit c4fa6c3c4eaacefd9bf50b5c88c2c72474a15bc5 (HEAD -> bb-10.4-release-igor, origin/bb-10.4-release-igor) Author: Igor Babaev <igor(a)askmonty.org> Date: Tue Jan 31 13:14:53 2023 -0800 MDEV-30218 Incorrect optimization for rowid_filtering Correction over the last patch for this MDEV. I modify the code to add a printout: diff --git a/sql/sql_select.cc b/sql/sql_select.cc index 4bdfb659513..2cd4495bc03 100644 --- a/sql/sql_select.cc +++ b/sql/sql_select.cc @@ -7965,6 +7965,8 @@ best_access_path(JOIN *join, table->best_range_rowid_filter_for_partial_join(start_key->key, rows, access_cost_factor); + trace_access_idx.add("ROWS_FOR_FILTER", rows); + trace_access_idx.add("REAL_ROWS_FOR_FILTER", tmp); if (filter) { tmp-= filter->get_adjusted_gain(rows) - filter->get_cmp_gain(rows); then, I'm trying this example: create table t1 (a int, b int); insert into t1 select seq, seq from seq_1_to_10000; create table t2 (a int, key(a)); insert into t2 select seq from seq_1_to_10000; insert into t2 select * from t2; insert into t2 select * from t2; analyze table t2 persistent for all; create table t3 ( pk int primary key, a int, b int, unique key(a), key(b) ); insert into t3 select seq, seq, seq from seq_1_to_100000; set optimizer_trace=1; explain select straight_join * from t1, t2, t3 where t2.a=t1.a and t3.a=t1.b and t3.b < 5000; +------+-------------+-------+---------------+---------------+------+---------+---------+--------+---------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +------+-------------+-------+---------------+---------------+------+---------+---------+--------+---------------------------------+ | 1 | SIMPLE | t1 | ALL | NULL | NULL | NULL | NULL | 10157 | Using where | | 1 | SIMPLE | t2 | ref | a | a | 5 | j1.t1.a | 4 | Using index | | 1 | SIMPLE | t3 | eq_ref|filter | a,b | a|b | 5|5 | j1.t1.b | 1 (5%) | Using where; Using rowid filter | +------+-------------+-------+---------------+---------------+------+---------+---------+--------+---------------------------------+ good so far. select * from information_schema.optimizer_trace; shows this (note the added printouts in caps): { "plan_prefix": ["t1", "t2"], "table": "t3", "best_access_path": { "considered_access_paths": [ { "access_type": "eq_ref", "index": "a", "ROWS_FOR_FILTER": 40628, "REAL_ROWS_FOR_FILTER": 10157, "rowid_filter_key": "b", "rows": 1, "cost": 9498, "chosen": true }, Let's run the query and check what really happens: ANALYZE FORMAT=JSON select straight_join * from t1, t2, t3 where t2.a=t1.a and t3.a=t1.b and t3.b < 5000; shows: ... "table": { "table_name": "t3", "access_type": "eq_ref", "key": "a", ... "rowid_filter": { ... "r_lookups": 10000, ... }, "r_loops": 40000, ... So, ROWS_FOR_FILTER=40K REAL_ROWS_FOR_FILTER=10K which of the numbers is a closer match here for the observed value of r_lookups= 10K ? BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net

1 0

[Maria-developers] Unable to reset my mariadb root password.
by sanaboyina sarvani 01 Feb '23

01 Feb '23

Hello, I have just installed mariadb in my macos system. After installation it asks for the root password which I have not set yet. Tried to reset the root password but that did not work. So, how to know my mariadb root password? Is it possible to reset the password? Thank you. Regards

1 0

[Maria-developers] Optimizer call followup: negative gains from rowid filters
by Sergey Petrunia 31 Jan '23

31 Jan '23

Hello, Just taking notes from the optimizer call: Check out this branch: 10.4-mdev30218-print-negative-gain (git log -3 output below) It has: - 10.4-release contents - then mdev30218_bb-10.4-release.diff patch provided by Igor - then a debug patch by me. I run: ./mtr --mem --parallel=5 --force main/rowid_filter*test grep -rn filter_gain var/ I get this: negative gains from filters: var/4/log/mysqld.1.err:52:filter_gain:-0.0569276, query=select 1 var/4/log/mysqld.1.err:58:filter_gain:-0.0569276, query=explain extended select 1 var/4/log/mysqld.1.err:64:filter_gain:-0.250816, query=SELECT * FROM t1 HAVING (7, 9) IN (SELECT t2.i1, t2.i2 FROM t2 WHERE t2.i1 = 3) var/4/log/mysqld.1.err:65:filter_gain:-0.250816, query=EXPLAIN EXTENDED var/4/log/mysqld.1.err:67:filter_gain:-10.207, query=explain var/4/log/mysqld.1.err:69:filter_gain:-10.207, query=SELECT * FROM t1 WHERE a > 0 AND b=0 var/4/log/mysqld.1.err:70:filter_gain:-1.17502, query=select count(*) from t1 where a between 21 and 30 and b=2 var/4/log/mysqld.1.err:71:filter_gain:-1.17502, query=explain extended select count(*) from t1 where a between 21 and 30 and b=2 var/4/log/mysqld.1.err:72:filter_gain:-1.17502, query=select * from t1 where a between 21 and 30 and b=2 var/4/log/mysqld.1.err:73:filter_gain:-2.18259, query=SELECT * FROM t1 var/4/log/mysqld.1.err:77:filter_gain:-2.18259, query=EXPLAIN EXTENDED SELECT * FROM t1 var/4/log/mysqld.1.err:81:filter_gain:-0.151308, query=set statement optimizer_switch='rowid_filter=on' for explain extended select t.id, fi.* var/4/log/mysqld.1.err:90:filter_gain:-0.151308, query=set statement optimizer_switch='rowid_filter=on' for select t.id, fi.* var/4/log/mysqld.1.err:99:filter_gain:-0.151308, query=set statement optimizer_switch='rowid_filter=on' for explain extended select t.id, fi.* var/4/log/mysqld.1.err:108:filter_gain:-0.151308, query=set statement optimizer_switch='rowid_filter=on' for select t.id, fi.* var/4/log/mysqld.1.err:117:filter_gain:-0.151308, query=set statement optimizer_switch='rowid_filter=on' for analyze format=json select t.id, fi.* var/3/log/mysqld.1.err:51:filter_gain:-0.0569276, query=select 1 var/3/log/mysqld.1.err:57:filter_gain:-0.0569276, query=explain extended select 1 var/3/log/mysqld.1.err:63:filter_gain:-0.250816, query=SELECT * FROM t1 HAVING (7, 9) IN (SELECT t2.i1, t2.i2 FROM t2 WHERE t2.i1 = 3) var/3/log/mysqld.1.err:64:filter_gain:-0.250816, query=EXPLAIN EXTENDED var/3/log/mysqld.1.err:66:filter_gain:-12.7213, query=explain var/3/log/mysqld.1.err:68:filter_gain:-12.7213, query=SELECT * FROM t1 WHERE a > 0 AND b=0 === git log -3 output === commit 82d5977b9fd6be5f5152d6831cd744fc739a2f24 (HEAD -> 10.4-mdev30218-print-negative-gain, origin/10.4-mdev30218-print-negative-gain, bb-10.4-release) Author: Sergei Petrunia <sergey(a)mariadb.com> Date: Tue Jan 31 20:11:17 2023 +0300 Debug patch: print filter_gain if it is less than 0. commit 47a40d90627dcf3c870cecd33c345c2e224986af Author: Sergei Petrunia <sergey(a)mariadb.com> Date: Tue Jan 31 20:07:02 2023 +0300 Apply mdev30218_bb-10.4-release.diff commit c8f2e9a5c0ac5905f28b050b7df5a9ffd914b7e7 (origin/bb-10.4-release) Author: Oleksandr Byelkin <sanja(a)mariadb.com> Date: Mon Jan 30 11:49:42 2023 +0100 Fix number of rows passing in case of EQ_REF BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://petrunia.net

1 0

Re: [Maria-developers] dcd69cbd19a: MDEV-29447 MDEV-26285 Refactor spider_db_mbase_util::open_item_func
by Sergei Golubchik 30 Jan '23

30 Jan '23

Hi, Yuchen, Thanks! There are two kinds of comments below, you'll see which is which. I realize that this patch mainly just moves parts of spider code around. Some of my comments apply to the new code that didn't exist before and appeared only in this patch. It's what needs to be fixed (or discussed, if you disagree) before pushing. Other comments apply to the old spider code that this patch moved. If you'd like you can say "I'm just moving spider code around" and ignore all comments of that kind. /Sergei On Jan 22, Yuchen Pei wrote: > commit dcd69cbd19a > Author: Yuchen Pei <yuchen.pei(a)mariadb.com> > Date: Tue Jan 3 16:24:04 2023 +1100 > > MDEV-29447 MDEV-26285 Refactor spider_db_mbase_util::open_item_func > > Porting commit 3836098c29ef1b7ff9d5fbde99b690eab73a0df1 (MDEV-26285) Because 3836098c29ef will not be, apparently, pushed anywhere, you'd better put a self-containing comment here. May be use the comment of 3836098c29ef? Note that you shouldn't push your 10.4-11.0 versions of the patch directly, only push the 10.3 one (when ok) and let it propagate to 11.0 through merges. Your 10.4-11.0 patches are very useful, though, they will serve as guidelines for whoever will merge this change up. > to current versions 10.3+ to fix a problem (MDEV-29447) where field > items that are arguments of a func item may be used before created / > initialised. > > Signed-off-by: Yuchen Pei <yuchen.pei(a)mariadb.com> > > diff --git a/storage/spider/spd_db_mysql.cc b/storage/spider/spd_db_mysql.cc > index e942d1d9063..a886a4e6e99 100644 > --- a/storage/spider/spd_db_mysql.cc > +++ b/storage/spider/spd_db_mysql.cc > @@ -4035,6 +4042,170 @@ int spider_db_mbase_util::open_item_func( > uint alias_length, > bool use_fields, > spider_fields *fields > +) { > + DBUG_ENTER("spider_db_mbase_util::check_item_func"); incorrect function name in DBUG_ENTER > + > + int error = check_item_func(item_func, spider, alias, > + alias_length, use_fields, fields); > + if (error) > + DBUG_RETURN(error); > + if (!str) > + DBUG_RETURN(0); > + > + DBUG_RETURN(print_item_func(item_func, spider, str, alias, > + alias_length, use_fields, fields)); > +} > + > +namespace { > + bool item_func_is_timestampdiff( We generally use static in these cases. Why would an anonymous namespace be better? > + const char *func_name, > + int func_name_length > + ) { > + return func_name_length == 13 && > + !strncasecmp("timestampdiff", func_name, func_name_length); We build with rtti nowadays, so you can use dynamic_cast instead. and you won't need func_name and func_name_length. Ideally you won't need a check for timestampdiff at all > + } > + > + bool not_func_should_be_skipped( > + Item_func *item_func > + ){ > + DBUG_ENTER("not_func_should_be_skipped"); add DBUG_ASSERT(item_func->functype() == Item_func::NOT_FUNC); > + Item **item_list = item_func->arguments(); > + > + if (item_list[0]->type() == Item::COND_ITEM) > + { > + DBUG_PRINT("info",("spider item_list[0] is COND_ITEM")); > + Item_cond *item_cond = (Item_cond *) item_list[0]; > + if (item_cond->functype() == Item_func::COND_AND_FUNC) > + { > + DBUG_PRINT("info",("spider item_cond is COND_AND_FUNC")); > + List_iterator_fast<Item> lif(*(item_cond->argument_list())); > + bool has_expr_cache_item = FALSE; > + bool has_isnotnull_func = FALSE; > + bool has_other_item = FALSE; > + while(Item *item = lif++) > + { > + if ( > + item->type() == Item::EXPR_CACHE_ITEM > + ) { > + DBUG_PRINT("info",("spider EXPR_CACHE_ITEM")); > + has_expr_cache_item = TRUE; > + } else > + if ( > + item->type() == Item::FUNC_ITEM && > + ((Item_func *) item)->functype() == Item_func::ISNOTNULL_FUNC > + ) { > + DBUG_PRINT("info",("spider ISNOTNULL_FUNC")); > + has_isnotnull_func = TRUE; > + } else { > + DBUG_PRINT("info",("spider has other item")); > + DBUG_PRINT("info",("spider COND type=%d", item->type())); > + has_other_item = TRUE; > + } > + } > + if (has_expr_cache_item && has_isnotnull_func && !has_other_item) So, "not func should be skipped" if it's NOT (expr1 AND expr2 AND ...) and all expr's are either IS NOT NULL or a cached subquery. Why? IS NULL, IS NOT FALSE, non-cached subqueries, everything else is fine, IS NOT NULL if it is followed by `AND 2>1` is fine, OR is fine. Doesn't make much sense to me. > + { > + DBUG_PRINT("info",("spider NOT EXISTS skip")); > + DBUG_RETURN(TRUE); > + } > + } > + } > + DBUG_RETURN(FALSE); > + } > +} > + > +/** > + Check if the given item_func and its arguments can be pushed down to > + a data node. This function is recursive because we need to also check > + the arguments of the item_func. > + > + @return 0: success, otherwise: error > + */ > +int spider_db_mbase_util::check_item_func( > + Item_func *item_func, > + ha_spider *spider, > + const char *alias, > + uint alias_length, > + bool use_fields, > + spider_fields *fields > +) { > + DBUG_ENTER("spider_db_mbase_util::check_item_func"); > + > + Item_func::Functype func_type = item_func->functype(); > + DBUG_PRINT("info",("spider functype = %d", func_type)); > + > + const char *func_name = (char*) item_func->func_name(); > + int func_name_length = strlen(func_name); > + DBUG_PRINT("info",("spider func_name = %s", func_name)); > + > + /* The blacklist of the functions that cannot be pushed down */ > + if ( > + func_type == Item_func::TRIG_COND_FUNC || > + func_type == Item_func::CASE_SEARCHED_FUNC || > + func_type == Item_func::CASE_SIMPLE_FUNC Does CASE need to be skipped at all? > + ) { > + DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > + } > + > + if (func_type == Item_func::NOT_FUNC) > + { > + /* Why the following check is necessary? */ indeed > + if(not_func_should_be_skipped(item_func)) > + DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > + } > + > + if (func_type == Item_func::UDF_FUNC) old code used switch here. I think it looks more natural than a sequence of if's. A compiler might optimize them internally into a switch, but I'd better do it explicitly. > + { > + int use_pushdown_udf = spider_param_use_pushdown_udf( > + spider->trx->thd, spider->share->use_pushdown_udf); > + if (!use_pushdown_udf) { > + DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > + } > + } > + > + if (func_type == Item_func::FT_FUNC) { > + if (spider_db_check_ft_idx(item_func, spider) == MAX_KEY) > + DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > + } > + > + if (func_type == Item_func::UNKNOWN_FUNC) > + { > + if (item_func_is_timestampdiff(func_name, func_name_length)) { it doesn't seem that it needs to be skipped at all > + DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > + } > + } > + /* End of the blacklist */ > + > + /* Check the arguments recursively */ > + if (uint item_count = item_func->argument_count()) > + { > + Item **item_list = item_func->arguments(); > + for (uint roop_count = 0; roop_count < item_count; roop_count++) > + { > + Item *item = item_list[roop_count]; > + if (int error_num = spider_db_print_item_type(item, NULL, spider, NULL, > + alias, alias_length, dbton_id, use_fields, fields)) it'd be more logical to invoke check_item_func() here instead of spider_db_print_item_type with str==NULL > + DBUG_RETURN(error_num); > + } > + } > + > + DBUG_RETURN(0); > +} > + > +/** > + The function print the string corresponding to the given item_func to str. > + The function is assumed to be called only when the check by the function > + check_item_func() has passed. > + > + @return 0: success, otherwise: error > + */ > +int spider_db_mbase_util::print_item_func( > + Item_func *item_func, > + ha_spider *spider, > + spider_string *str, > + const char *alias, > + uint alias_length, > + bool use_fields, > + spider_fields *fields > ) { > int error_num; > Item *item, **item_list = item_func->arguments(); > @@ -4049,13 +4220,14 @@ int spider_db_mbase_util::open_item_func( > last_str_length = SPIDER_SQL_NULL_CHAR_LEN; > int use_pushdown_udf; > bool merge_func = FALSE; > - DBUG_ENTER("spider_db_mbase_util::open_item_func"); > - if (str) > - { > + DBUG_ENTER("spider_db_mbase_util::print_item_func"); > + DBUG_ASSERT(!check_item_func(item_func, spider, alias, alias_length, > + use_fields, fields) && str); Make it two asserts please. One for !check_item_func(...) and the other DBUG_ASSERT(str); Generally, always try to split assert into as many separate asserts as possible, don't do assert(A && B); So that when it fails it'd provide as much information as possible about what exactly went wrong. > + > if (str->reserve(SPIDER_SQL_OPEN_PAREN_LEN)) > DBUG_RETURN(HA_ERR_OUT_OF_MEM); > str->q_append(SPIDER_SQL_OPEN_PAREN_STR, SPIDER_SQL_OPEN_PAREN_LEN); > - } > + > DBUG_PRINT("info",("spider functype = %d", item_func->functype())); > switch (item_func->functype()) > { > @@ -4374,94 +4523,12 @@ int spider_db_mbase_util::open_item_func( > { > if (!strncasecmp("utc_timestamp", func_name, func_name_length)) pretty much all that code can use dynamic_cast nowadays > { > - if (str) > str->length(str->length() - SPIDER_SQL_OPEN_PAREN_LEN); > DBUG_RETURN(spider_db_open_item_string(item_func, NULL, spider, str, > alias, alias_length, dbton_id, use_fields, fields)); > } else if (!strncasecmp("timestampdiff", func_name, func_name_length)) > { > -#ifdef ITEM_FUNC_TIMESTAMPDIFF_ARE_PUBLIC but it _is_ public, so you can push down Item_func_timestamp_diff all right > - Item_func_timestamp_diff *item_func_timestamp_diff = > - (Item_func_timestamp_diff *) item_func; > - if (str) > - { > - const char *interval_str; > - uint interval_len; > - switch (item_func_timestamp_diff->int_type) > - { > - case INTERVAL_YEAR: > - interval_str = SPIDER_SQL_YEAR_STR; > - interval_len = SPIDER_SQL_YEAR_LEN; > - break; > - case INTERVAL_QUARTER: > - interval_str = SPIDER_SQL_QUARTER_STR; > - interval_len = SPIDER_SQL_QUARTER_LEN; > - break; > - case INTERVAL_MONTH: > - interval_str = SPIDER_SQL_MONTH_STR; > - interval_len = SPIDER_SQL_MONTH_LEN; > - break; > - case INTERVAL_WEEK: > - interval_str = SPIDER_SQL_WEEK_STR; > - interval_len = SPIDER_SQL_WEEK_LEN; > - break; > - case INTERVAL_DAY: > - interval_str = SPIDER_SQL_DAY_STR; > - interval_len = SPIDER_SQL_DAY_LEN; > - break; > - case INTERVAL_HOUR: > - interval_str = SPIDER_SQL_HOUR_STR; > - interval_len = SPIDER_SQL_HOUR_LEN; > - break; > - case INTERVAL_MINUTE: > - interval_str = SPIDER_SQL_MINUTE_STR; > - interval_len = SPIDER_SQL_MINUTE_LEN; > - break; > - case INTERVAL_SECOND: > - interval_str = SPIDER_SQL_SECOND_STR; > - interval_len = SPIDER_SQL_SECOND_LEN; > - break; > - case INTERVAL_MICROSECOND: > - interval_str = SPIDER_SQL_MICROSECOND_STR; > - interval_len = SPIDER_SQL_MICROSECOND_LEN; > - break; > - default: > - interval_str = ""; > - interval_len = 0; > - break; > - } > - str->length(str->length() - SPIDER_SQL_OPEN_PAREN_LEN); > - if (str->reserve(func_name_length + SPIDER_SQL_OPEN_PAREN_LEN + > - interval_len + SPIDER_SQL_COMMA_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(func_name, func_name_length); > - str->q_append(SPIDER_SQL_OPEN_PAREN_STR, SPIDER_SQL_OPEN_PAREN_LEN); > - str->q_append(interval_str, interval_len); > - str->q_append(SPIDER_SQL_COMMA_STR, SPIDER_SQL_COMMA_LEN); > - } > - if ((error_num = spider_db_print_item_type(item_list[0], NULL, spider, > - str, alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - if (str) > - { > - if (str->reserve(SPIDER_SQL_COMMA_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_COMMA_STR, SPIDER_SQL_COMMA_LEN); > - } > - if ((error_num = spider_db_print_item_type(item_list[1], NULL, spider, > - str, alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - if (str) > - { > - if (str->reserve(SPIDER_SQL_CLOSE_PAREN_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_CLOSE_PAREN_STR, > - SPIDER_SQL_CLOSE_PAREN_LEN); > - } > - DBUG_RETURN(0); > -#else > DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > -#endif > } > } else if (func_name_length == 14) > { > @@ -5038,84 +4996,17 @@ int spider_db_mbase_util::open_item_func( > func_name = (char*)item_func->func_name(); > func_name_length = strlen(func_name); > } > - } > break; > case Item_func::CASE_SEARCHED_FUNC: > case Item_func::CASE_SIMPLE_FUNC: > -#ifdef ITEM_FUNC_CASE_PARAMS_ARE_PUBLIC Item_func_case looks somewhat different today, but I suspect it can be pushed too > - Item_func_case *item_func_case = (Item_func_case *) item_func; > - if (str) > - { > - if (str->reserve(SPIDER_SQL_CASE_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_CASE_STR, SPIDER_SQL_CASE_LEN); > - } > - if (item_func_case->first_expr_num != -1) > - { > - if ((error_num = spider_db_print_item_type( > - item_list[item_func_case->first_expr_num], NULL, spider, str, > - alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - } > - for (roop_count = 0; roop_count < item_func_case->ncases; > - roop_count += 2) > - { > - if (str) > - { > - if (str->reserve(SPIDER_SQL_WHEN_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_WHEN_STR, SPIDER_SQL_WHEN_LEN); > - } > - if ((error_num = spider_db_print_item_type( > - item_list[roop_count], NULL, spider, str, > - alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - if (str) > - { > - if (str->reserve(SPIDER_SQL_THEN_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_THEN_STR, SPIDER_SQL_THEN_LEN); > - } > - if ((error_num = spider_db_print_item_type( > - item_list[roop_count + 1], NULL, spider, str, > - alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - } > - if (item_func_case->else_expr_num != -1) > - { > - if (str) > - { > - if (str->reserve(SPIDER_SQL_ELSE_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_ELSE_STR, SPIDER_SQL_ELSE_LEN); > - } > - if ((error_num = spider_db_print_item_type( > - item_list[item_func_case->else_expr_num], NULL, spider, str, > - alias, alias_length, dbton_id, use_fields, fields))) > - DBUG_RETURN(error_num); > - } > - if (str) > - { > - if (str->reserve(SPIDER_SQL_END_LEN + SPIDER_SQL_CLOSE_PAREN_LEN)) > - DBUG_RETURN(HA_ERR_OUT_OF_MEM); > - str->q_append(SPIDER_SQL_END_STR, SPIDER_SQL_END_LEN); > - str->q_append(SPIDER_SQL_CLOSE_PAREN_STR, > - SPIDER_SQL_CLOSE_PAREN_LEN); > - } > - DBUG_RETURN(0); > -#else > DBUG_RETURN(ER_SPIDER_COND_SKIP_NUM); > -#endif > case Item_func::JSON_EXTRACT_FUNC: > func_name = (char*) item_func->func_name(); > func_name_length = strlen(func_name); > - if (str) > - { > if (str->reserve(func_name_length + SPIDER_SQL_OPEN_PAREN_LEN)) > DBUG_RETURN(HA_ERR_OUT_OF_MEM); > str->q_append(func_name, func_name_length); > str->q_append(SPIDER_SQL_OPEN_PAREN_STR, SPIDER_SQL_OPEN_PAREN_LEN); > - } > func_name = SPIDER_SQL_COMMA_STR; > func_name_length = SPIDER_SQL_COMMA_LEN; > separator_str = SPIDER_SQL_COMMA_STR; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

2 1

[Maria-developers] To getting started for contribution at your GitHub repo
by Omkar shivaji Devgirikar 29 Jan '23

29 Jan '23

Respect sir, I Omkar Devgirikar , I am a student and as a beginner i want to start contributing in your server repo with C++ language . Can you please guide me with your project like resources for this project, how get started and some good first issues. Please look at my mail seriously and guide me please. I hope you will look at it. Thank You

2 2