1berry/MariaDB-server
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge 10.5 into 10.6

Browse Source
This commit is contained in:
Marko Mäkelä 2021-05-18 11:13:45 +03:00
commit f09d33f521
31 changed files with 606 additions and 329 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

111
client/mysqlbinlog.cc
View File

@ -114,7 +114,7 @@ static void warning(const char *format, ...) ATTRIBUTE_FORMAT(printf, 1, 2);
static bool one_database=0, one_table=0, to_last_remote_log= 0, disable_log_bin= 0;
static bool opt_hexdump= 0, opt_version= 0;
const char *base64_output_mode_names[]=
{"NEVER", "AUTO", "ALWAYS", "UNSPEC", "DECODE-ROWS", NullS};
{"NEVER", "AUTO", "UNSPEC", "DECODE-ROWS", NullS};
TYPELIB base64_output_mode_typelib=
{ array_elements(base64_output_mode_names) - 1, "",
base64_output_mode_names, NULL };
@ -836,53 +836,6 @@ static bool shall_skip_table(const char *log_tblname)
strcmp(log_tblname, table);
}
/**
Prints the given event in base64 format.
The header is printed to the head cache and the body is printed to
the body cache of the print_event_info structure. This allows all
base64 events corresponding to the same statement to be joined into
one BINLOG statement.
@param[in] ev Log_event to print.
@param[in,out] result_file FILE to which the output will be written.
@param[in,out] print_event_info Parameters and context state
determining how to print.
@retval ERROR_STOP An error occurred - the program should terminate.
@retval OK_CONTINUE No error, the program should continue.
*/
static Exit_status
write_event_header_and_base64(Log_event *ev, FILE *result_file,
PRINT_EVENT_INFO *print_event_info)
{
IO_CACHE *head= &print_event_info->head_cache;
IO_CACHE *body= &print_event_info->body_cache;
DBUG_ENTER("write_event_header_and_base64");
/* Write header and base64 output to cache */
if (ev->print_header(head, print_event_info, FALSE))
DBUG_RETURN(ERROR_STOP);
DBUG_ASSERT(print_event_info->base64_output_mode == BASE64_OUTPUT_ALWAYS);
if (ev->print_base64(body, print_event_info,
print_event_info->base64_output_mode !=
BASE64_OUTPUT_DECODE_ROWS))
DBUG_RETURN(ERROR_STOP);
/* Read data from cache and write to result file */
if (copy_event_cache_to_file_and_reinit(head, result_file) ||
copy_event_cache_to_file_and_reinit(body, result_file))
{
error("Error writing event to file.");
DBUG_RETURN(ERROR_STOP);
}
DBUG_RETURN(OK_CONTINUE);
}
static bool print_base64(PRINT_EVENT_INFO *print_event_info, Log_event *ev)
{
/*
@ -1134,19 +1087,9 @@ Exit_status process_event(PRINT_EVENT_INFO *print_event_info, Log_event *ev,
qe->flags|= LOG_EVENT_SUPPRESS_USE_F;
}
print_use_stmt(print_event_info, qe);
if (opt_base64_output_mode == BASE64_OUTPUT_ALWAYS)
{
if ((retval= write_event_header_and_base64(ev, result_file,
print_event_info)) !=
OK_CONTINUE)
goto end;
}
else
{
print_skip_replication_statement(print_event_info, ev);
if (ev->print(result_file, print_event_info))
goto err;
}
print_skip_replication_statement(print_event_info, ev);
if (ev->print(result_file, print_event_info))
goto err;
if (head->error == -1)
goto err;
break;
@ -1170,19 +1113,9 @@ Exit_status process_event(PRINT_EVENT_INFO *print_event_info, Log_event *ev,
filename and use LOCAL), prepared in the 'case EXEC_LOAD_EVENT'
below.
*/
if (opt_base64_output_mode == BASE64_OUTPUT_ALWAYS)
{
if ((retval= write_event_header_and_base64(ce, result_file,
print_event_info)) !=
OK_CONTINUE)
goto end;
}
else
{
print_skip_replication_statement(print_event_info, ev);
if (ce->print(result_file, print_event_info, TRUE))
goto err;
}
print_skip_replication_statement(print_event_info, ev);
if (ce->print(result_file, print_event_info, TRUE))
goto err;
// If this binlog is not 3.23 ; why this test??
if (glob_description_event->binlog_version >= 3)
{
@ -1589,12 +1522,10 @@ static struct my_option my_options[] =
"--verbose option is also given; "
"'auto' prints base64 only when necessary (i.e., for row-based events and "
"format description events); "
"'always' prints base64 whenever possible. "
"--base64-output with no 'name' argument is equivalent to "
"--base64-output=always and is also deprecated. If no "
"--base64-output[=name] option is given at all, the default is 'auto'.",
"If no --base64-output=name option is given at all, the default is "
"'auto'.",
&opt_base64_output_mode_str, &opt_base64_output_mode_str,
0, GET_STR, OPT_ARG, 0, 0, 0, 0, 0, 0},
0, GET_STR, REQUIRED_ARG, 0, 0, 0, 0, 0, 0},
/*
mysqlbinlog needs charsets knowledge, to be able to convert a charset
number found in binlog to a charset name (to be able to print things
@ -2038,20 +1969,15 @@ get_one_option(const struct my_option *opt, const char *argument, const char *fi
stop_datetime= convert_str_to_timestamp(stop_datetime_str);
break;
case OPT_BASE64_OUTPUT_MODE:
if (argument == NULL)
opt_base64_output_mode= BASE64_OUTPUT_ALWAYS;
else
{
int val;
int val;
if ((val= find_type_with_warning(argument, &base64_output_mode_typelib,
opt->name)) <= 0)
{
sf_leaking_memory= 1; /* no memory leak reports here */
die();
}
opt_base64_output_mode= (enum_base64_output_mode) (val - 1);
if ((val= find_type_with_warning(argument, &base64_output_mode_typelib,
opt->name)) <= 0)
{
sf_leaking_memory= 1; /* no memory leak reports here */
die();
}
opt_base64_output_mode= (enum_base64_output_mode)(val - 1);
break;
case OPT_REWRITE_DB: // db_from->db_to
{
@ -2901,8 +2827,7 @@ static Exit_status check_header(IO_CACHE* file,
(ulonglong)tmp_pos);
return ERROR_STOP;
}
if (opt_base64_output_mode == BASE64_OUTPUT_AUTO
|| opt_base64_output_mode == BASE64_OUTPUT_ALWAYS)
if (opt_base64_output_mode == BASE64_OUTPUT_AUTO)
{
/*
process_event will delete *description_event and set it to

18
man/mysqlbinlog.1
View File

@ -141,7 +141,7 @@ Display a help message and exit\&.
.\}
.\" mysqlbinlog: base64-output option
.\" base64-output option: mysqlbinlog
\fB\-\-base64\-output[=\fR\fB\fIvalue\fR\fR\fB]\fR
\fB\-\-base64\-output=\fR\fB\fIvalue\fR\fR\fB\fR
.sp
This option determines when events should be displayed encoded as base\-64 strings using
BINLOG
@ -192,22 +192,6 @@ to re\-execute binary log file contents\&. The other option values are intended
.sp -1
.IP \(bu 2.3
.\}
ALWAYS
displays
BINLOG
statements whenever possible\&. This is the implied value if the option is given as
\fB\-\-base64\-output\fR
without a value\&. Both ALWAYS and not giving a value are deprecated.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
NEVER
causes
BINLOG

2
mysql-test/include/commit.inc
View File

@ -775,7 +775,7 @@ call p_verify_status_increment(2, 0, 2, 0);
commit;
call p_verify_status_increment(0, 0, 0, 0);
check table t1, t2, t3;
call p_verify_status_increment(6, 0, 6, 0);
call p_verify_status_increment(4, 0, 4, 0);
commit;
call p_verify_status_increment(0, 0, 0, 0);
drop view v1;

2
mysql-test/lib/My/Debugger.pm
View File

@ -139,7 +139,7 @@ sub do_args($$$$$) {
my $v = $debuggers{$k};
# on windows mtr args are quoted (for system), otherwise not (for exec)
sub quote($) { $_[0] =~ / / ? "\"$_[0]\"" : $_[0] }
sub quote($) { $_[0] =~ /[; ]/ ? "\"$_[0]\"" : $_[0] }
sub unquote($) { $_[0] =~ s/^"(.*)"$/$1/; $_[0] }
sub quote_from_mtr($) { IS_WINDOWS() ? $_[0] : quote($_[0]) }
sub unquote_for_mtr($) { IS_WINDOWS() ? $_[0] : unquote($_[0]) }

2
mysql-test/main/commit_1innodb.result
View File

@ -879,7 +879,7 @@ Table Op Msg_type Msg_text
test.t1 check status OK
test.t2 check status OK
test.t3 check status OK
call p_verify_status_increment(6, 0, 6, 0);
call p_verify_status_increment(4, 0, 4, 0);
SUCCESS
commit;

15
mysql-test/main/mysqlbinlog.result
View File

@ -878,8 +878,19 @@ ROLLBACK /* added by mysqlbinlog */;
/*!50530 SET @@SESSION.PSEUDO_SLAVE_MODE=0*/;
End of 5.0 tests
End of 5.1 tests
# Expect deprecation warning.
# Expect deprecation warning again.
#
# Expect error for invalid --base64-output argument value.
# MYSQL_BINLOG std_data/master-bin.000001 --base64-output=always 2>&1
Unknown option to base64-output: always
Alternatives are: 'NEVER','AUTO','UNSPEC','DECODE-ROWS'
#
# Expect error for incomplete --base64-output argument.
# MYSQL_BINLOG std_data/master-bin.000001 --base64-output 2>&1
mysqlbinlog: option '--base64-output' requires an argument
#
# Ensure --base64-output=auto outputs the same result as unspecified
# MYSQL_BINLOG -v MYSQLD_DATADIR/master-bin.000001 > MYSQLTEST_VARDIR/tmp/mysqlbinlog_nob64spec.out
# MYSQL_BINLOG --base64-output=auto -v MYSQLD_DATADIR/master-bin.000001 > MYSQLTEST_VARDIR/tmp/mysqlbinlog_b64auto.out
RESET MASTER;
CREATE DATABASE test1;
USE test1;

32
mysql-test/main/mysqlbinlog.test
View File

@ -521,18 +521,32 @@ remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn.empty;
remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn;
#
# WL#5625: Deprecate mysqlbinlog options --base64-output=always and --base64-output
# MDEV-25222: Remove mysqlbinlog options --base64-output=always and --base64-output
#
--echo # Expect deprecation warning.
--exec $MYSQL_BINLOG --base64-output=always std_data/master-bin.000001 > /dev/null 2> $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn
--cat_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn
--echo #
--echo # Expect error for invalid --base64-output argument value.
--echo # MYSQL_BINLOG std_data/master-bin.000001 --base64-output=always 2>&1
--error 1
--exec $MYSQL_BINLOG std_data/master-bin.000001 --base64-output=always 2>&1
--echo # Expect deprecation warning again.
--exec $MYSQL_BINLOG --base64-output std_data/master-bin.000001 > /dev/null 2> $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn
--cat_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn
--echo #
--echo # Expect error for incomplete --base64-output argument.
--echo # MYSQL_BINLOG std_data/master-bin.000001 --base64-output 2>&1
# The error produces the absolute path of the mysqlbinlog executable, remove it.
--replace_regex /.*mysqlbinlog.*:/mysqlbinlog:/i
--error 1
--exec $MYSQL_BINLOG std_data/master-bin.000001 --base64-output 2>&1
--echo #
--echo # Ensure --base64-output=auto outputs the same result as unspecified
--echo # MYSQL_BINLOG -v MYSQLD_DATADIR/master-bin.000001 > MYSQLTEST_VARDIR/tmp/mysqlbinlog_nob64spec.out
--exec $MYSQL_BINLOG -v $MYSQLD_DATADIR/master-bin.000001 > $MYSQLTEST_VARDIR/tmp/mysqlbinlog_nob64spec.out
--echo # MYSQL_BINLOG --base64-output=auto -v MYSQLD_DATADIR/master-bin.000001 > MYSQLTEST_VARDIR/tmp/mysqlbinlog_b64auto.out
--exec $MYSQL_BINLOG --base64-output=auto -v $MYSQLD_DATADIR/master-bin.000001 > $MYSQLTEST_VARDIR/tmp/mysqlbinlog_b64auto.out
--diff_files $MYSQLTEST_VARDIR/tmp/mysqlbinlog_nob64spec.out $MYSQLTEST_VARDIR/tmp/mysqlbinlog_b64auto.out
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog_nob64spec.out
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog_b64auto.out
# Clean up this part of the test.
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.warn
# BUG#50914
# This test verifies if the approach of the mysqlbinlog prints

40
mysql-test/suite/binlog/r/binlog_admin_cmd_kill.result Normal file
View File

@ -0,0 +1,40 @@
#
# Kill OPTIMIZE command prior to table modification
#
RESET MASTER;
CREATE TABLE t1 (f INT) ENGINE=INNODB;
CREATE TABLE t2 (f INT) ENGINE=INNODB;
connect con1,127.0.0.1,root,,test,$MASTER_MYPORT,;
connection con1;
SET debug_sync='admin_command_kill_before_modify SIGNAL ready_to_be_killed WAIT_FOR master_cont';
OPTIMIZE TABLE t1,t2;
connection default;
SET debug_sync='now WAIT_FOR ready_to_be_killed';
KILL THD_ID;
SET debug_sync = 'reset';
disconnect con1;
include/show_binlog_events.inc
Log_name Pos Event_type Server_id End_log_pos Info
master-bin.000001 # Gtid # # GTID #-#-#
master-bin.000001 # Query # # use `test`; CREATE TABLE t1 (f INT) ENGINE=INNODB
master-bin.000001 # Gtid # # GTID #-#-#
master-bin.000001 # Query # # use `test`; CREATE TABLE t2 (f INT) ENGINE=INNODB
DROP TABLE t1,t2;
FLUSH LOGS;
#
# Kill OPTIMIZE command after table modification
#
CREATE TABLE t1 (f INT) ENGINE=INNODB;
CREATE TABLE t2 (f INT) ENGINE=INNODB;
connect con1,127.0.0.1,root,,test,$MASTER_MYPORT,;
connection con1;
SET debug_sync='admin_command_kill_after_modify SIGNAL ready_to_be_killed WAIT_FOR master_cont';
OPTIMIZE TABLE t1,t2;
connection default;
SET debug_sync='now WAIT_FOR ready_to_be_killed';
KILL THD_ID;
SET debug_sync = 'reset';
disconnect con1;
DROP TABLE t1,t2;
FLUSH LOGS;
FOUND 1 /OPTIMIZE TABLE t1,t2/ in mysqlbinlog.out

95
mysql-test/suite/binlog/t/binlog_admin_cmd_kill.test Normal file
View File

@ -0,0 +1,95 @@
# ==== Purpose ====
#
# Test verifies that when an admin command execution is interrupted by KILL
# command it should stop its execution. The admin command in binary log should
# contain only the list of tables which have successfully executed admin
# command prior to kill.
#
# ==== Implementation ====
#
# Steps:
# 0 - Create two table t1,t2.
# 1 - Execute OPTIMIZE TABLE t1,t2 command.
# 2 - Using debug sync mechanism kill OPTIMIZE TABLE command at a stage
# where it has not optimized any table.
# 3 - Check that OPTIMIZE TABLE command is not written to binary log.
# 4 - Using debug sync mechanism hold the execution of OPTIMIZE TABLE after
# t1 table optimization. Now kill the OPTIMIZE TABLE command.
# 5 - Observe the binlog output, the OPTIMIZE TABLE command should display `t1,t2`.
# 6 - Please note that, we binlog the entire query even if at least one
# table is modified as admin commands are safe to replicate and they will
# not make the slave to diverge.
#
# ==== References ====
#
# MDEV-22530: Aborting OPTIMIZE TABLE still logs in binary log and replicates to the Slave server.
#
--source include/have_log_bin.inc
--source include/have_debug.inc
--source include/have_debug_sync.inc
--source include/have_innodb.inc
--echo #
--echo # Kill OPTIMIZE command prior to table modification
--echo #
RESET MASTER;
CREATE TABLE t1 (f INT) ENGINE=INNODB;
CREATE TABLE t2 (f INT) ENGINE=INNODB;
--connect(con1,127.0.0.1,root,,test,$MASTER_MYPORT,)
--connection con1
SET debug_sync='admin_command_kill_before_modify SIGNAL ready_to_be_killed WAIT_FOR master_cont';
--send OPTIMIZE TABLE t1,t2
--connection default
SET debug_sync='now WAIT_FOR ready_to_be_killed';
--let $thd_id= `SELECT ID FROM INFORMATION_SCHEMA.PROCESSLIST WHERE INFO LIKE '%OPTIMIZE TABLE %'`
# Now kill.
--replace_result $thd_id THD_ID
eval KILL $thd_id;
SET debug_sync = 'reset';
--disconnect con1
--source include/show_binlog_events.inc
DROP TABLE t1,t2;
FLUSH LOGS;
--echo #
--echo # Kill OPTIMIZE command after table modification
--echo #
CREATE TABLE t1 (f INT) ENGINE=INNODB;
CREATE TABLE t2 (f INT) ENGINE=INNODB;
--connect(con1,127.0.0.1,root,,test,$MASTER_MYPORT,)
--connection con1
SET debug_sync='admin_command_kill_after_modify SIGNAL ready_to_be_killed WAIT_FOR master_cont';
--send OPTIMIZE TABLE t1,t2
--connection default
SET debug_sync='now WAIT_FOR ready_to_be_killed';
--let $thd_id= `SELECT ID FROM INFORMATION_SCHEMA.PROCESSLIST WHERE INFO LIKE '%OPTIMIZE TABLE %'`
# Now kill.
--replace_result $thd_id THD_ID
eval KILL $thd_id;
SET debug_sync = 'reset';
--disconnect con1
DROP TABLE t1,t2;
let $binlog_file= query_get_value(SHOW MASTER STATUS, File, 1);
FLUSH LOGS;
--let $MYSQLD_DATADIR= `select @@datadir`
--exec $MYSQL_BINLOG $MYSQLD_DATADIR/$binlog_file > $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--let SEARCH_PATTERN= OPTIMIZE TABLE t1,t2
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out

1
mysql-test/suite/galera_3nodes/disabled.def
View File

@ -16,7 +16,6 @@ galera_gtid_2_cluster : MDEV-23775 Galera test failure on galera_3nodes.galera_g
galera_ist_gcache_rollover : MDEV-23578 WSREP: exception caused by message: {v=0,t=1,ut=255,o=4,s=0,sr=0,as=1,f=6,src=50524cfe,srcvid=view_id(REG,50524cfe,4),insvid=view_id(UNKNOWN,00000000,0),ru=00000000,r=[-1,-1],fs=75,nl=(}
galera_load_data_ist : MDEV-24639 galera_3nodes.galera_load_data_ist MTR failed with SIGABRT: query 'reap' failed: 2013: Lost connection to server during query
galera_load_data_ist : MDEV-24639 galera_3nodes.galera_load_data_ist MTR failed with SIGABRT: query 'reap' failed: 2013: Lost connection to server during query
galera_pc_bootstrap : MDEV-24650 galera_pc_bootstrap MTR failed: Could not execute 'check-testcase' before testcase
galera_safe_to_bootstrap : MDEV-24097 galera_3nodes.galera_safe_to_bootstrap MTR sporadaically fails: Failed to start mysqld or mysql_shutdown failed
galera_slave_options_do : MDEV-8798
galera_slave_options_ignore : MDEV-8798

79
mysql-test/suite/rpl/r/rpl_mark_optimize_tbl_ddl.result Normal file
View File

@ -0,0 +1,79 @@
include/rpl_init.inc [topology=1->2]
connection server_1;
FLUSH TABLES;
ALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB;
connection server_2;
SET @save_slave_parallel_threads= @@GLOBAL.slave_parallel_threads;
SET @save_slave_parallel_mode= @@GLOBAL.slave_parallel_mode;
include/stop_slave.inc
SET GLOBAL slave_parallel_threads=2;
SET GLOBAL slave_parallel_mode=optimistic;
include/start_slave.inc
connection server_1;
CREATE TABLE t1(a INT) ENGINE=INNODB;
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize note Table does not support optimize, doing recreate + analyze instead
test.t1 optimize status OK
INSERT INTO t1 VALUES(1);
INSERT INTO t1 SELECT 1+a FROM t1;
INSERT INTO t1 SELECT 2+a FROM t1;
connection server_2;
#
# Verify that following admin commands are marked as ddl
# 'OPTIMIZE TABLE', 'REPAIR TABLE' and 'ANALYZE TABLE'
#
connection server_1;
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize note Table does not support optimize, doing recreate + analyze instead
test.t1 optimize status OK
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair note The storage engine for the table doesn't support repair
ANALYZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 analyze status Engine-independent statistics collected
test.t1 analyze status OK
FLUSH LOGS;
FOUND 1 /GTID 0-1-8 ddl/ in mysqlbinlog.out
FOUND 1 /GTID 0-1-9 ddl/ in mysqlbinlog.out
FOUND 1 /GTID 0-1-10 ddl/ in mysqlbinlog.out
#
# Clean up
#
DROP TABLE t1;
connection server_2;
FLUSH LOGS;
#
# Check that ALTER TABLE commands with ANALYZE, OPTIMIZE and REPAIR on
# partitions will be marked as DDL in binary log.
#
connection server_1;
CREATE TABLE t1(id INT) PARTITION BY RANGE (id) (PARTITION p0 VALUES LESS THAN (100),
PARTITION pmax VALUES LESS THAN (MAXVALUE));
INSERT INTO t1 VALUES (1), (10), (100), (1000);
ALTER TABLE t1 ANALYZE PARTITION p0;
Table Op Msg_type Msg_text
test.t1 analyze status Engine-independent statistics collected
test.t1 analyze status OK
ALTER TABLE t1 OPTIMIZE PARTITION p0;
Table Op Msg_type Msg_text
test.t1 optimize status OK
ALTER TABLE t1 REPAIR PARTITION p0;
Table Op Msg_type Msg_text
test.t1 repair status OK
FLUSH LOGS;
FOUND 1 /GTID 0-1-14 ddl/ in mysqlbinlog.out
FOUND 1 /GTID 0-1-15 ddl/ in mysqlbinlog.out
FOUND 1 /GTID 0-1-16 ddl/ in mysqlbinlog.out
#
# Clean up
#
DROP TABLE t1;
connection server_2;
include/stop_slave.inc
SET GLOBAL slave_parallel_threads= @save_slave_parallel_threads;
SET GLOBAL slave_parallel_mode= @save_slave_parallel_mode;
include/start_slave.inc
include/rpl_end.inc

142
mysql-test/suite/rpl/t/rpl_mark_optimize_tbl_ddl.test Normal file
View File

@ -0,0 +1,142 @@
# ==== Purpose ====
#
# Test verifies that there is no deadlock or assertion in
# slave_parallel_mode=optimistic configuration while applying admin command
# like 'OPTIMIZE TABLE', 'REPAIR TABLE' and 'ANALYZE TABLE'.
#
# ==== Implementation ====
#
# Steps:
# 0 - Create a table, execute OPTIMIZE TABLE command on the table followed
# by some DMLS.
# 1 - No assert should happen on slave server.
# 2 - Assert that 'OPTIMIZE TABLE', 'REPAIR TABLE' and 'ANALYZE TABLE' are
# marked as 'DDL' in the binary log.
#
# ==== References ====
#
# MDEV-17515: GTID Replication in optimistic mode deadlock
#
--source include/have_partition.inc
--source include/have_innodb.inc
--let $rpl_topology=1->2
--source include/rpl_init.inc
--connection server_1
FLUSH TABLES;
ALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB;
--connection server_2
SET @save_slave_parallel_threads= @@GLOBAL.slave_parallel_threads;
SET @save_slave_parallel_mode= @@GLOBAL.slave_parallel_mode;
--source include/stop_slave.inc
SET GLOBAL slave_parallel_threads=2;
SET GLOBAL slave_parallel_mode=optimistic;
--source include/start_slave.inc
--connection server_1
CREATE TABLE t1(a INT) ENGINE=INNODB;
OPTIMIZE TABLE t1;
INSERT INTO t1 VALUES(1);
INSERT INTO t1 SELECT 1+a FROM t1;
INSERT INTO t1 SELECT 2+a FROM t1;
--save_master_pos
--connection server_2
--sync_with_master
--echo #
--echo # Verify that following admin commands are marked as ddl
--echo # 'OPTIMIZE TABLE', 'REPAIR TABLE' and 'ANALYZE TABLE'
--echo #
--connection server_1
OPTIMIZE TABLE t1;
--let optimize_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
REPAIR TABLE t1;
--let repair_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
ANALYZE TABLE t1;
--let analyze_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
let $binlog_file= query_get_value(SHOW MASTER STATUS, File, 1);
FLUSH LOGS;
--let $MYSQLD_DATADIR= `select @@datadir`
--exec $MYSQL_BINLOG $MYSQLD_DATADIR/$binlog_file > $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--let SEARCH_PATTERN= GTID $optimize_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--let SEARCH_PATTERN= GTID $repair_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--let SEARCH_PATTERN= GTID $analyze_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--echo #
--echo # Clean up
--echo #
DROP TABLE t1;
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--save_master_pos
--connection server_2
--sync_with_master
FLUSH LOGS;
--echo #
--echo # Check that ALTER TABLE commands with ANALYZE, OPTIMIZE and REPAIR on
--echo # partitions will be marked as DDL in binary log.
--echo #
--connection server_1
CREATE TABLE t1(id INT) PARTITION BY RANGE (id) (PARTITION p0 VALUES LESS THAN (100),
PARTITION pmax VALUES LESS THAN (MAXVALUE));
INSERT INTO t1 VALUES (1), (10), (100), (1000);
ALTER TABLE t1 ANALYZE PARTITION p0;
--let analyze_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
ALTER TABLE t1 OPTIMIZE PARTITION p0;
--let optimize_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
ALTER TABLE t1 REPAIR PARTITION p0;
--let repair_gtid= `SELECT @@GLOBAL.gtid_binlog_pos`
let $binlog_file= query_get_value(SHOW MASTER STATUS, File, 1);
FLUSH LOGS;
--exec $MYSQL_BINLOG $MYSQLD_DATADIR/$binlog_file > $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--let SEARCH_PATTERN= GTID $analyze_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--let SEARCH_PATTERN= GTID $optimize_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--let SEARCH_PATTERN= GTID $repair_gtid ddl
--let SEARCH_FILE= $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--source include/search_pattern_in_file.inc
--echo #
--echo # Clean up
--echo #
DROP TABLE t1;
--remove_file $MYSQLTEST_VARDIR/tmp/mysqlbinlog.out
--save_master_pos
--connection server_2
--sync_with_master
--source include/stop_slave.inc
SET GLOBAL slave_parallel_threads= @save_slave_parallel_threads;
SET GLOBAL slave_parallel_mode= @save_slave_parallel_mode;
--source include/start_slave.inc
--source include/rpl_end.inc

8
mysql-test/suite/rpl/t/rpl_slave_shutdown_mdev20821.test
View File

@ -4,6 +4,10 @@
# hang when the parallel workers were idle.
# The bug reported scenario is extented to cover the multi-sources case as well as
# checking is done for both the idle and busy workers cases.
#
# MDEV-25336 Parallel replication causes failed assert while restarting
# Since this test case involves slave restart this will help in testing
# Mdev-25336 too.
--source include/have_innodb.inc
--source include/have_binlog_format_mixed.inc
@ -26,7 +30,7 @@ select @@global.slave_parallel_workers as two;
# At this point worker threads have no assignement.
# Shutdown must not hang.
# In 10.2/10.3 there should not be any assert failure `prev != 0 && next != 0'
--connection server_3
--write_file $MYSQLTEST_VARDIR/tmp/mysqld.3.expect
wait
@ -75,6 +79,7 @@ insert into t1 values (1);
--connection server_3
--sync_with_master 0,''
# In 10.2/10.3 there should not be any assert failure `prev != 0 && next != 0'
# At this point worker threads have no assignement.
# Shutdown must not hang.
@ -117,6 +122,7 @@ insert into t1 values (2);
insert into t2 values (2);
# In 10.2/10.3 there should not be any assert failure `prev != 0 && next != 0'
# At this point there's a good chance the worker threads are busy.
# SHUTDOWN must proceed without any delay as above.
--connection server_3

12
sql/handler.h
View File

@ -1876,9 +1876,19 @@ struct THD_TRANS
CREATED_TEMP_TABLE= 2,
DROPPED_TEMP_TABLE= 4,
DID_WAIT= 8,
DID_DDL= 0x10
DID_DDL= 0x10,
EXECUTED_TABLE_ADMIN_CMD= 0x20
};
void mark_executed_table_admin_cmd()
{
DBUG_PRINT("debug", ("mark_executed_table_admin_cmd"));
m_unsafe_rollback_flags|= EXECUTED_TABLE_ADMIN_CMD;
}
bool trans_executed_admin_cmd()
{
return (m_unsafe_rollback_flags & EXECUTED_TABLE_ADMIN_CMD) != 0;
}
void mark_created_temp_table()
{
DBUG_PRINT("debug", ("mark_created_temp_table"));

5
sql/log_event.h
View File

@ -822,9 +822,8 @@ bool copy_event_cache_to_file_and_reinit(IO_CACHE *cache, FILE *file);
enum enum_base64_output_mode {
BASE64_OUTPUT_NEVER= 0,
BASE64_OUTPUT_AUTO= 1,
BASE64_OUTPUT_ALWAYS= 2,
BASE64_OUTPUT_UNSPEC= 3,
BASE64_OUTPUT_DECODE_ROWS= 4,
BASE64_OUTPUT_UNSPEC= 2,
BASE64_OUTPUT_DECODE_ROWS= 3,
/* insert new output modes here */
BASE64_OUTPUT_MODE_COUNT
};

4
sql/log_event_server.cc
View File

@ -3262,8 +3262,10 @@ Gtid_log_event::Gtid_log_event(THD *thd_arg, uint64 seq_no_arg,
flags2|= FL_WAITED;
if (thd_arg->transaction->stmt.trans_did_ddl() ||
thd_arg->transaction->stmt.has_created_dropped_temp_table() ||
thd_arg->transaction->stmt.trans_executed_admin_cmd() ||
thd_arg->transaction->all.trans_did_ddl() ||
thd_arg->transaction->all.has_created_dropped_temp_table())
thd_arg->transaction->all.has_created_dropped_temp_table() ||
thd_arg->transaction->all.trans_executed_admin_cmd())
flags2|= FL_DDL;
else if (is_transactional && !is_tmp_table)
flags2|= FL_TRANSACTIONAL;

4
sql/rpl_rli.cc
View File

@ -210,8 +210,8 @@ a file name for --relay-log-index option", opt_relaylog_index_name);
*/
sql_print_warning("Neither --relay-log nor --relay-log-index were used;"
" so replication "
"may break when this MySQL server acts as a "
"slave and has his hostname changed!! Please "
"may break when this MariaDB server acts as a "
"replica and has its hostname changed. Please "
"use '--log-basename=#' or '--relay-log=%s' to avoid "
"this problem.", ln);
name_warning_sent= 1;

77
sql/sql_admin.cc
View File

@ -1,5 +1,5 @@
/* Copyright (c) 2010, 2015, Oracle and/or its affiliates.
Copyright (c) 2011, 2020, MariaDB
Copyright (c) 2011, 2021, MariaDB
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
@ -490,7 +490,8 @@ static bool mysql_admin_table(THD* thd, TABLE_LIST* tables,
int (handler::*operator_func)(THD *,
HA_CHECK_OPT *),
int (view_operator_func)(THD *, TABLE_LIST*,
HA_CHECK_OPT *))
HA_CHECK_OPT *),
bool is_cmd_replicated)
{
TABLE_LIST *table;
List<Item> field_list;
@ -501,6 +502,8 @@ static bool mysql_admin_table(THD* thd, TABLE_LIST* tables,
int compl_result_code;
bool need_repair_or_alter= 0;
wait_for_commit* suspended_wfc;
bool is_table_modified= false;
DBUG_ENTER("mysql_admin_table");
DBUG_PRINT("enter", ("extra_open_options: %u", extra_open_options));
@ -560,6 +563,10 @@ static bool mysql_admin_table(THD* thd, TABLE_LIST* tables,
bool open_for_modify= org_open_for_modify;
DBUG_PRINT("admin", ("table: '%s'.'%s'", db, table->table_name.str));
DEBUG_SYNC(thd, "admin_command_kill_before_modify");
if (thd->is_killed())
break;
strxmov(table_name, db, ".", table->table_name.str, NullS);
thd->open_options|= extra_open_options;
table->lock_type= lock_type;
@ -1212,6 +1219,13 @@ send_result_message:
break;
}
}
/*
Admin commands acquire table locks and these locks are not detected by
parallel replication deadlock detection-and-handling mechanism. Hence
they must be marked as DDL so that they are not scheduled in parallel
with conflicting DMLs resulting in deadlock.
*/
thd->transaction->stmt.mark_executed_table_admin_cmd();
if (table->table && !table->view)
{
/*
@ -1253,10 +1267,9 @@ send_result_message:
}
else
{
if (trans_commit_stmt(thd) ||
(stmt_causes_implicit_commit(thd, CF_IMPLICIT_COMMIT_END) &&
trans_commit_implicit(thd)))
if (trans_commit_stmt(thd))
goto err;
is_table_modified= true;
}
close_thread_tables(thd);
thd->release_transactional_locks();
@ -1279,6 +1292,16 @@ send_result_message:
if (protocol->write())
goto err;
DEBUG_SYNC(thd, "admin_command_kill_after_modify");
}
if (is_table_modified && is_cmd_replicated &&
(!opt_readonly || thd->slave_thread) && !thd->lex->no_write_to_binlog)
{
thd->get_stmt_da()->set_overwrite_status(true);
auto res= write_bin_log(thd, true, thd->query(), thd->query_length());
thd->get_stmt_da()->set_overwrite_status(false);
if (res)
goto err;
}
my_eof(thd);
@ -1341,7 +1364,7 @@ bool mysql_assign_to_keycache(THD* thd, TABLE_LIST* tables,
check_opt.key_cache= key_cache;
DBUG_RETURN(mysql_admin_table(thd, tables, &check_opt,
"assign_to_keycache", TL_READ_NO_INSERT, 0, 0,
0, 0, &handler::assign_to_keycache, 0));
0, 0, &handler::assign_to_keycache, 0, false));
}
@ -1368,7 +1391,7 @@ bool mysql_preload_keys(THD* thd, TABLE_LIST* tables)
*/
DBUG_RETURN(mysql_admin_table(thd, tables, 0,
"preload_keys", TL_READ_NO_INSERT, 0, 0, 0, 0,
&handler::preload_keys, 0));
&handler::preload_keys, 0, false));
}
@ -1389,17 +1412,8 @@ bool Sql_cmd_analyze_table::execute(THD *thd)
WSREP_TO_ISOLATION_BEGIN_WRTCHK(NULL, NULL, first_table);
res= mysql_admin_table(thd, first_table, &m_lex->check_opt,
"analyze", lock_type, 1, 0, 0, 0,
&handler::ha_analyze, 0);
/* ! we write after unlocking the table */
if (!res && !m_lex->no_write_to_binlog && (!opt_readonly || thd->slave_thread))
{
/*
Presumably, ANALYZE and binlog writing doesn't require synchronization
*/
thd->get_stmt_da()->set_overwrite_status(true);
res= write_bin_log(thd, TRUE, thd->query(), thd->query_length());
thd->get_stmt_da()->set_overwrite_status(false);
}
&handler::ha_analyze, 0, true);
m_lex->first_select_lex()->table_list.first= first_table;
m_lex->query_tables= first_table;
@ -1425,7 +1439,7 @@ bool Sql_cmd_check_table::execute(THD *thd)
res= mysql_admin_table(thd, first_table, &m_lex->check_opt, "check",
lock_type, 0, 0, HA_OPEN_FOR_REPAIR, 0,
&handler::ha_check, &view_check);
&handler::ha_check, &view_check, false);
m_lex->first_select_lex()->table_list.first= first_table;
m_lex->query_tables= first_table;
@ -1450,17 +1464,8 @@ bool Sql_cmd_optimize_table::execute(THD *thd)
mysql_recreate_table(thd, first_table, true) :
mysql_admin_table(thd, first_table, &m_lex->check_opt,
"optimize", TL_WRITE, 1, 0, 0, 0,
&handler::ha_optimize, 0);
/* ! we write after unlocking the table */
if (!res && !m_lex->no_write_to_binlog && (!opt_readonly || thd->slave_thread))
{
/*
Presumably, OPTIMIZE and binlog writing doesn't require synchronization
*/
thd->get_stmt_da()->set_overwrite_status(true);
res= write_bin_log(thd, TRUE, thd->query(), thd->query_length());
thd->get_stmt_da()->set_overwrite_status(false);
}
&handler::ha_optimize, 0, true);
m_lex->first_select_lex()->table_list.first= first_table;
m_lex->query_tables= first_table;
@ -1483,18 +1488,8 @@ bool Sql_cmd_repair_table::execute(THD *thd)
TL_WRITE, 1,
MY_TEST(m_lex->check_opt.sql_flags & TT_USEFRM),
HA_OPEN_FOR_REPAIR, &prepare_for_repair,
&handler::ha_repair, &view_repair);
&handler::ha_repair, &view_repair, true);
/* ! we write after unlocking the table */
if (!res && !m_lex->no_write_to_binlog && (!opt_readonly || thd->slave_thread))
{
/*
Presumably, REPAIR and binlog writing doesn't require synchronization
*/
thd->get_stmt_da()->set_overwrite_status(true);
res= write_bin_log(thd, TRUE, thd->query(), thd->query_length());
thd->get_stmt_da()->set_overwrite_status(false);
}
m_lex->first_select_lex()->table_list.first= first_table;
m_lex->query_tables= first_table;

3
sql/sql_class.h
View File

@ -5382,7 +5382,8 @@ public:
transaction->all.m_unsafe_rollback_flags|=
(transaction->stmt.m_unsafe_rollback_flags &
(THD_TRANS::DID_WAIT | THD_TRANS::CREATED_TEMP_TABLE |
THD_TRANS::DROPPED_TEMP_TABLE | THD_TRANS::DID_DDL));
THD_TRANS::DROPPED_TEMP_TABLE | THD_TRANS::DID_DDL |
THD_TRANS::EXECUTED_TABLE_ADMIN_CMD));
}

1
sql/sql_prepare.cc
View File

@ -6098,6 +6098,7 @@ extern "C" int execute_sql_command(const char *command,
new_thd->store_globals();
new_thd->security_ctx->skip_grants();
new_thd->query_cache_is_applicable= 0;
new_thd->variables.wsrep_on= 0;
bzero((char*) &new_thd->net, sizeof(new_thd->net));
thd= new_thd;
}

54
storage/innobase/handler/ha_innodb.cc
View File

@ -3866,7 +3866,7 @@ innobase_commit_low(
if (trx_is_started(trx)) {
trx_commit_for_mysql(trx);
} else {
trx->will_lock = 0;
trx->will_lock = false;
#ifdef WITH_WSREP
trx->wsrep = false;
#endif /* WITH_WSREP */
@ -7360,7 +7360,7 @@ ha_innobase::write_row(
}
if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
ins_mode_t vers_set_fields;
@ -8116,7 +8116,7 @@ ha_innobase::update_row(
if (is_read_only()) {
DBUG_RETURN(HA_ERR_TABLE_READONLY);
} else if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
if (m_upd_buf == NULL) {
@ -8284,7 +8284,7 @@ ha_innobase::delete_row(
if (is_read_only()) {
DBUG_RETURN(HA_ERR_TABLE_READONLY);
} else if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
if (!m_prebuilt->upd_node) {
@ -8554,8 +8554,12 @@ ha_innobase::index_read(
/* For R-Tree index, we will always place the page lock to
pages being searched */
if (dict_index_is_spatial(index)) {
++m_prebuilt->trx->will_lock;
if (index->is_spatial() && !m_prebuilt->trx->will_lock) {
if (trx_is_started(m_prebuilt->trx)) {
DBUG_RETURN(HA_ERR_READ_ONLY_TRANSACTION);
} else {
m_prebuilt->trx->will_lock = true;
}
}
/* Note that if the index for which the search template is built is not
@ -9117,7 +9121,7 @@ ha_innobase::ft_init()
them as regular read only transactions for now. */
if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
DBUG_RETURN(rnd_init(false));
@ -9183,7 +9187,7 @@ ha_innobase::ft_init_ext(
them as regular read only transactions for now. */
if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
dict_table_t* ft_table = m_prebuilt->table;
@ -12707,8 +12711,7 @@ void
create_table_info_t::allocate_trx()
{
m_trx = innobase_trx_allocate(m_thd);
m_trx->will_lock = 1;
m_trx->will_lock = true;
}
/** Create a new table to an InnoDB database.
@ -13026,13 +13029,7 @@ inline int ha_innobase::delete_table(const char* name, enum_sql_command sqlcom)
ut_a(name_len < 1000);
/* Either the transaction is already flagged as a locking transaction
or it hasn't been started yet. */
ut_a(!trx_is_started(trx) || trx->will_lock > 0);
/* We are doing a DDL operation. */
++trx->will_lock;
trx->will_lock = true;
/* Drop the table in InnoDB */
@ -13206,14 +13203,7 @@ innobase_drop_database(
#endif /* _WIN32 */
trx_t* trx = innobase_trx_allocate(thd);
/* Either the transaction is already flagged as a locking transaction
or it hasn't been started yet. */
ut_a(!trx_is_started(trx) || trx->will_lock > 0);
/* We are doing a DDL operation. */
++trx->will_lock;
trx->will_lock = true;
ulint dummy;
@ -13250,7 +13240,7 @@ inline dberr_t innobase_rename_table(trx_t *trx, const char *from,
DEBUG_SYNC_C("innodb_rename_table_ready");
trx_start_if_not_started(trx, true);
ut_ad(trx->will_lock > 0);
ut_ad(trx->will_lock);
if (commit) {
/* Serialize data dictionary operations with dictionary mutex:
@ -13355,7 +13345,7 @@ int ha_innobase::truncate()
const char* name = mem_heap_strdup(heap, ib_table->name.m_name);
trx_t* trx = innobase_trx_allocate(m_user_thd);
trx->will_lock = 1;
trx->will_lock = true;
trx->dict_operation = true;
row_mysql_lock_data_dictionary(trx);
dict_stats_wait_bg_to_stop_using_table(ib_table, trx);
@ -13442,7 +13432,7 @@ ha_innobase::rename_table(
trx_t* trx = innobase_trx_allocate(thd);
/* We are doing a DDL operation. */
trx->will_lock = 1;
trx->will_lock = true;
trx->dict_operation = true;
dberr_t error = innobase_rename_table(trx, from, to, true);
@ -15316,7 +15306,7 @@ ha_innobase::start_stmt(
innobase_register_trx(ht, thd, trx);
if (!trx_is_started(trx)) {
++trx->will_lock;
trx->will_lock = true;
}
DBUG_RETURN(0);
@ -15561,7 +15551,7 @@ ha_innobase::external_lock(
&& (m_prebuilt->select_lock_type != LOCK_NONE
|| m_prebuilt->stored_select_lock_type != LOCK_NONE)) {
++trx->will_lock;
trx->will_lock = true;
}
DBUG_RETURN(0);
@ -15600,7 +15590,7 @@ ha_innobase::external_lock(
&& (m_prebuilt->select_lock_type != LOCK_NONE
|| m_prebuilt->stored_select_lock_type != LOCK_NONE)) {
++trx->will_lock;
trx->will_lock = true;
}
DBUG_RETURN(0);
@ -15946,7 +15936,7 @@ ha_innobase::store_lock(
&& (m_prebuilt->select_lock_type != LOCK_NONE
|| m_prebuilt->stored_select_lock_type != LOCK_NONE)) {
++trx->will_lock;
trx->will_lock = true;
}
return(to);

3
storage/innobase/handler/handler0alter.cc
View File

@ -2360,7 +2360,7 @@ innodb_instant_alter_column_allowed_reason:
}
}
m_prebuilt->trx->will_lock++;
m_prebuilt->trx->will_lock = true;
/* When changing a NULL column to NOT NULL and specifying a
DEFAULT value, ensure that the DEFAULT expression is a constant.
@ -11382,7 +11382,6 @@ foreign_fail:
m_prebuilt = ctx->prebuilt;
}
trx_start_if_not_started(user_trx, true);
user_trx->will_lock++;
m_prebuilt->trx = user_trx;
}
DBUG_INJECT_CRASH("ib_commit_inplace_crash",

2
storage/innobase/include/trx0i_s.h
View File

@ -150,7 +150,7 @@ struct i_s_trx_row_t {
bool trx_is_read_only;
/*!< trx_t::read_only */
bool trx_is_autocommit_non_locking;
/*!< trx_is_autocommit_non_locking(trx)
/*!< trx:t::is_autocommit_non_locking()
*/
};

2
storage/innobase/include/trx0sys.h
View File

@ -514,7 +514,7 @@ class rw_trx_hash_t
static void validate_element(trx_t *trx)
{
ut_ad(!trx->read_only || !trx->rsegs.m_redo.rseg);
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
/* trx->state can be anything except TRX_STATE_NOT_STARTED */
ut_d(trx->mutex_lock());
ut_ad(trx_state_eq(trx, TRX_STATE_ACTIVE) ||

68
storage/innobase/include/trx0trx.h
View File

@ -332,64 +332,6 @@ from innodb_lock_wait_timeout via trx_t::mysql_thd.
? thd_lock_wait_timeout((t)->mysql_thd) \
: 0)
/**
Determine if the transaction is a non-locking autocommit select
(implied read-only).
@param t transaction
@return true if non-locking autocommit select transaction. */
#define trx_is_autocommit_non_locking(t) \
((t)->auto_commit && (t)->will_lock == 0)
/**
Determine if the transaction is a non-locking autocommit select
with an explicit check for the read-only status.
@param t transaction
@return true if non-locking autocommit read-only transaction. */
#define trx_is_ac_nl_ro(t) \
((t)->read_only && trx_is_autocommit_non_locking((t)))
/**
Check transaction state */
#define check_trx_state(t) do { \
ut_ad(!trx_is_autocommit_non_locking((t))); \
switch ((t)->state) { \
case TRX_STATE_PREPARED: \
case TRX_STATE_PREPARED_RECOVERED: \
case TRX_STATE_ACTIVE: \
case TRX_STATE_COMMITTED_IN_MEMORY: \
continue; \
case TRX_STATE_NOT_STARTED: \
break; \
} \
ut_error; \
} while (0)
#ifdef UNIV_DEBUG
/*******************************************************************//**
Assert that an autocommit non-locking select cannot be in the
rw_trx_hash and that it is a read-only transaction.
The transaction must have mysql_thd assigned. */
# define assert_trx_nonlocking_or_in_list(t) \
do { \
if (trx_is_autocommit_non_locking(t)) { \
trx_state_t t_state = (t)->state; \
ut_ad((t)->read_only); \
ut_ad(!(t)->is_recovered); \
ut_ad((t)->mysql_thd); \
ut_ad(t_state == TRX_STATE_NOT_STARTED \
|| t_state == TRX_STATE_ACTIVE); \
} else { \
check_trx_state(t); \
} \
} while (0)
#else /* UNIV_DEBUG */
/*******************************************************************//**
Assert that an autocommit non-locking slect cannot be in the
rw_trx_hash and that it is a read-only transaction.
The transaction must have mysql_thd assigned. */
# define assert_trx_nonlocking_or_in_list(trx) ((void)0)
#endif /* UNIV_DEBUG */
typedef std::vector<ib_lock_t*, ut_allocator<ib_lock_t*> > lock_list;
/** The locks and state of an active transaction. Protected by
@ -882,16 +824,15 @@ public:
/*------------------------------*/
bool read_only; /*!< true if transaction is flagged
as a READ-ONLY transaction.
if auto_commit && will_lock == 0
if auto_commit && !will_lock
then it will be handled as a
AC-NL-RO-SELECT (Auto Commit Non-Locking
Read Only Select). A read only
transaction will not be assigned an
UNDO log. */
bool auto_commit; /*!< true if it is an autocommit */
ib_uint32_t will_lock; /*!< Will acquire some locks. Increment
each time we determine that a lock will
be acquired by the MySQL layer. */
bool will_lock; /*!< set to inform trx_start_low() that
the transaction may acquire locks */
/*------------------------------*/
fts_trx_t* fts_trx; /*!< FTS information, or NULL if
transaction hasn't modified tables
@ -1062,6 +1003,9 @@ public:
it->second.end_bulk_insert();
}
/** @return whether this is a non-locking autocommit transaction */
bool is_autocommit_non_locking() const { return auto_commit && !will_lock; }
/** This has to be invoked on SAVEPOINT or at the start of a statement.
Even if TRX_UNDO_EMPTY records were written for any table to cover an
insert into an empty table, subsequent operations will have to be covered

8
storage/innobase/include/trx0trx.ic
View File

@ -49,11 +49,15 @@ trx_state_eq(
case TRX_STATE_PREPARED:
case TRX_STATE_PREPARED_RECOVERED:
case TRX_STATE_COMMITTED_IN_MEMORY:
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
return(trx->state == state);
case TRX_STATE_ACTIVE:
assert_trx_nonlocking_or_in_list(trx);
if (trx->is_autocommit_non_locking()) {
ut_ad(!trx->is_recovered);
ut_ad(trx->read_only);
ut_ad(trx->mysql_thd);
}
return(state == trx->state);
case TRX_STATE_NOT_STARTED:

30
storage/innobase/lock/lock0lock.cc
View File

@ -1097,6 +1097,19 @@ static void lock_reset_lock_and_trx_wait(lock_t *lock)
lock->type_mode&= ~LOCK_WAIT;
}
#ifdef UNIV_DEBUG
/** Check transaction state */
static void check_trx_state(const trx_t *trx)
{
ut_ad(!trx->auto_commit || trx->will_lock);
const auto state= trx->state;
ut_ad(state == TRX_STATE_ACTIVE ||
state == TRX_STATE_PREPARED_RECOVERED ||
state == TRX_STATE_PREPARED ||
state == TRX_STATE_COMMITTED_IN_MEMORY);
}
#endif
/** Create a new record lock and inserts it to the lock queue,
without checking for deadlocks or conflicts.
@param[in] c_lock conflicting lock
@ -1127,7 +1140,7 @@ lock_rec_create_low(
ut_ad(dict_index_is_clust(index) || !dict_index_is_online_ddl(index));
ut_ad(!(type_mode & LOCK_TABLE));
ut_ad(trx->state != TRX_STATE_NOT_STARTED);
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
/* If rec is the supremum record, then we reset the gap and
LOCK_REC_NOT_GAP bits, as all locks on the supremum are
@ -3083,7 +3096,7 @@ lock_table_create(
ut_ad(trx->mutex_is_owner());
ut_ad(!trx->is_wsrep() || lock_sys.is_writer());
ut_ad(trx->state == TRX_STATE_ACTIVE || trx->is_recovered);
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
switch (LOCK_MODE_MASK & type_mode) {
case LOCK_AUTO_INC:
@ -4334,7 +4347,8 @@ lock_rec_queue_validate(
ut_ad(!index || lock->index == index);
lock->trx->mutex_lock();
ut_ad(!trx_is_ac_nl_ro(lock->trx));
ut_ad(!lock->trx->read_only
|| !lock->trx->is_autocommit_non_locking());
ut_ad(trx_state_eq(lock->trx,
TRX_STATE_COMMITTED_IN_MEMORY)
|| !lock->is_waiting()
@ -4420,8 +4434,8 @@ func_exit:
for (lock = lock_sys_t::get_first(cell, id, heap_no);
lock != NULL;
lock = lock_rec_get_next_const(heap_no, lock)) {
ut_ad(!trx_is_ac_nl_ro(lock->trx));
ut_ad(!lock->trx->read_only
|| !lock->trx->is_autocommit_non_locking());
ut_ad(!page_rec_is_metadata(rec));
if (index) {
@ -4497,7 +4511,8 @@ loop:
}
}
ut_ad(!trx_is_ac_nl_ro(lock->trx));
ut_ad(!lock->trx->read_only
|| !lock->trx->is_autocommit_non_locking());
/* Only validate the record queues when this thread is not
holding a tablespace latch. */
@ -4560,7 +4575,8 @@ lock_rec_validate(
lock != NULL;
lock = static_cast<const lock_t*>(HASH_GET_NEXT(hash, lock))) {
ut_ad(!trx_is_ac_nl_ro(lock->trx));
ut_ad(!lock->trx->read_only
|| !lock->trx->is_autocommit_non_locking());
ut_ad(!lock->is_table());
page_id_t current(lock->un_member.rec_lock.page_id);

4
storage/innobase/read/read0read.cc
View File

@ -1,7 +1,7 @@
/*****************************************************************************
Copyright (c) 1996, 2015, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2018, 2020, MariaDB Corporation.
Copyright (c) 2018, 2021, 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
@ -221,7 +221,7 @@ void ReadView::open(trx_t *trx)
else if (likely(!srv_read_only_mode))
{
m_creator_trx_id= trx->id;
if (trx_is_autocommit_non_locking(trx) && empty() &&
if (trx->is_autocommit_non_locking() && empty() &&
low_limit_id() == trx_sys.get_max_trx_id())
m_open.store(true, std::memory_order_relaxed);
else

21
storage/innobase/trx/trx0i_s.cc
View File

@ -521,7 +521,7 @@ thd_done:
row->trx_is_read_only = trx->read_only;
row->trx_is_autocommit_non_locking = trx_is_autocommit_non_locking(trx);
row->trx_is_autocommit_non_locking = trx->is_autocommit_non_locking();
return(TRUE);
}
@ -1165,7 +1165,24 @@ static void fetch_data_into_cache_low(trx_i_s_cache_t *cache, const trx_t *trx)
{
i_s_locks_row_t *requested_lock_row;
assert_trx_nonlocking_or_in_list(trx);
#ifdef UNIV_DEBUG
{
const auto state= trx->state;
if (trx->is_autocommit_non_locking())
{
ut_ad(trx->read_only);
ut_ad(!trx->is_recovered);
ut_ad(trx->mysql_thd);
ut_ad(state == TRX_STATE_NOT_STARTED || state == TRX_STATE_ACTIVE);
}
else
ut_ad(state == TRX_STATE_ACTIVE ||
state == TRX_STATE_PREPARED ||
state == TRX_STATE_PREPARED_RECOVERED ||
state == TRX_STATE_COMMITTED_IN_MEMORY);
}
#endif /* UNIV_DEBUG */
if (add_trx_relevant_locks_to_cache(cache, trx, &requested_lock_row))
{

32
storage/innobase/trx/trx0roll.cc
View File

@ -99,10 +99,21 @@ inline void trx_t::rollback_low(trx_savept_t *savept)
roll_node_t *roll_node= roll_node_create(heap);
roll_node->savept= savept;
if (savept)
check_trx_state(this);
else
assert_trx_nonlocking_or_in_list(this);
ut_ad(!in_rollback);
#ifdef UNIV_DEBUG
{
const auto s= state;
ut_ad(s == TRX_STATE_ACTIVE ||
s == TRX_STATE_PREPARED ||
s == TRX_STATE_PREPARED_RECOVERED);
if (savept)
{
ut_ad(s == TRX_STATE_ACTIVE);
ut_ad(mysql_thd);
ut_ad(!is_recovered);
}
}
#endif
error_state = DB_SUCCESS;
@ -198,7 +209,7 @@ dberr_t trx_rollback_for_mysql(trx_t* trx)
switch (trx->state) {
case TRX_STATE_NOT_STARTED:
trx->will_lock = 0;
trx->will_lock = false;
ut_ad(trx->mysql_thd);
#ifdef WITH_WSREP
trx->wsrep= false;
@ -208,12 +219,13 @@ dberr_t trx_rollback_for_mysql(trx_t* trx)
case TRX_STATE_ACTIVE:
ut_ad(trx->mysql_thd);
assert_trx_nonlocking_or_in_list(trx);
ut_ad(!trx->is_recovered);
ut_ad(!trx->is_autocommit_non_locking() || trx->read_only);
return(trx_rollback_for_mysql_low(trx));
case TRX_STATE_PREPARED:
case TRX_STATE_PREPARED_RECOVERED:
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
if (trx->rsegs.m_redo.undo || trx->rsegs.m_redo.old_insert) {
/* The XA ROLLBACK of a XA PREPARE transaction
will consist of multiple mini-transactions.
@ -260,7 +272,7 @@ dberr_t trx_rollback_for_mysql(trx_t* trx)
return(trx_rollback_for_mysql_low(trx));
case TRX_STATE_COMMITTED_IN_MEMORY:
check_trx_state(trx);
ut_ad(!trx->is_autocommit_non_locking());
break;
}
@ -289,7 +301,9 @@ trx_rollback_last_sql_stat_for_mysql(
return(DB_SUCCESS);
case TRX_STATE_ACTIVE:
assert_trx_nonlocking_or_in_list(trx);
ut_ad(trx->mysql_thd);
ut_ad(!trx->is_recovered);
ut_ad(!trx->is_autocommit_non_locking() || trx->read_only);
trx->op_info = "rollback of SQL statement";

58
storage/innobase/trx/trx0trx.cc
View File

@ -129,7 +129,7 @@ trx_init(
trx->auto_commit = false;
trx->will_lock = 0;
trx->will_lock = false;
trx->internal = false;
@ -342,7 +342,7 @@ trx_t *trx_create()
ib_alloc_t* alloc;
/* We just got trx from pool, it should be non locking */
ut_ad(trx->will_lock == 0);
ut_ad(!trx->will_lock);
ut_ad(!trx->rw_trx_hash_pins);
DBUG_LOG("trx", "Create: " << trx);
@ -542,7 +542,7 @@ void trx_disconnect_prepared(trx_t *trx)
trx->is_recovered= true;
trx->mysql_thd= NULL;
/* todo/fixme: suggest to do it at innodb prepare */
trx->will_lock= 0;
trx->will_lock= false;
trx_sys.rw_trx_hash.put_pins(trx);
}
@ -882,11 +882,10 @@ static trx_rseg_t* trx_assign_rseg_low()
/** Assign a rollback segment for modifying temporary tables.
@return the assigned rollback segment */
trx_rseg_t*
trx_t::assign_temp_rseg()
trx_rseg_t *trx_t::assign_temp_rseg()
{
ut_ad(!rsegs.m_noredo.rseg);
ut_ad(!trx_is_autocommit_non_locking(this));
ut_ad(!is_autocommit_non_locking());
compile_time_assert(ut_is_2pow(TRX_SYS_N_RSEGS));
/* Choose a temporary rollback segment between 0 and 127
@ -933,8 +932,8 @@ trx_start_low(
&& thd_trx_is_read_only(trx->mysql_thd));
if (!trx->auto_commit) {
++trx->will_lock;
} else if (trx->will_lock == 0) {
trx->will_lock = true;
} else if (!trx->will_lock) {
trx->read_only = true;
}
@ -972,7 +971,7 @@ trx_start_low(
trx_sys.register_rw(trx);
} else {
if (!trx_is_autocommit_non_locking(trx)) {
if (!trx->is_autocommit_non_locking()) {
/* If this is a read-only transaction that is writing
to a temporary table then it needs a transaction id
@ -1267,12 +1266,15 @@ inline void trx_t::commit_in_memory(const mtr_t *mtr)
must_flush_log_later= false;
read_view.close();
if (trx_is_autocommit_non_locking(this))
if (is_autocommit_non_locking())
{
ut_ad(id == 0);
ut_ad(read_only);
ut_ad(!will_lock);
ut_a(!is_recovered);
ut_ad(!rsegs.m_redo.rseg);
ut_ad(mysql_thd);
ut_ad(state == TRX_STATE_ACTIVE);
/* Note: We do not have to hold any lock_sys latch here, because
this is a non-locking transaction. */
@ -1284,12 +1286,11 @@ inline void trx_t::commit_in_memory(const mtr_t *mtr)
However, the freezing of trx_sys.trx_list will protect the trx_t
instance and it cannot be removed from the trx_list and freed
without first unfreezing trx_list. */
ut_ad(trx_state_eq(this, TRX_STATE_ACTIVE));
state= TRX_STATE_NOT_STARTED;
MONITOR_INC(MONITOR_TRX_NL_RO_COMMIT);
DBUG_LOG("trx", "Autocommit in memory: " << this);
state= TRX_STATE_NOT_STARTED;
}
else
{
@ -1439,8 +1440,6 @@ inline void trx_t::commit_in_memory(const mtr_t *mtr)
@param mtr mini-transaction (if there are any persistent modifications) */
void trx_t::commit_low(mtr_t *mtr)
{
assert_trx_nonlocking_or_in_list(this);
ut_ad(!trx_state_eq(this, TRX_STATE_COMMITTED_IN_MEMORY));
ut_ad(!mtr || mtr->is_active());
ut_d(bool aborted = in_rollback && error_state == DB_DEADLOCK);
ut_ad(!mtr == (aborted || !has_logged_or_recovered()));
@ -1449,13 +1448,13 @@ void trx_t::commit_low(mtr_t *mtr)
/* undo_no is non-zero if we're doing the final commit. */
if (fts_trx && undo_no)
{
ut_a(!trx_is_autocommit_non_locking(this));
dberr_t error= fts_commit(this);
ut_a(!is_autocommit_non_locking());
/* FTS-FIXME: Temporarily tolerate DB_DUPLICATE_KEY instead of
dying. This is a possible scenario if there is a crash between
insert to DELETED table committing and transaction committing. The
fix would be able to return error from this function */
ut_a(error == DB_SUCCESS || error == DB_DUPLICATE_KEY);
if (dberr_t error= fts_commit(this))
ut_a(error == DB_DUPLICATE_KEY);
}
#ifndef DBUG_OFF
@ -1700,9 +1699,6 @@ trx_print_low(
ulint heap_size)
/*!< in: mem_heap_get_size(trx->lock.lock_heap) */
{
ibool newline;
const char* op_info;
if (const trx_id_t id = trx->id) {
fprintf(f, "TRANSACTION " TRX_ID_FMT, trx->id);
} else {
@ -1729,11 +1725,7 @@ trx_print_low(
fprintf(f, ", state %lu", (ulong) trx->state);
ut_ad(0);
state_ok:
/* prevent a race condition */
op_info = trx->op_info;
if (*op_info) {
if (const char *op_info = trx->op_info) {
putc(' ', f);
fputs(op_info, f);
}
@ -1750,18 +1742,18 @@ state_ok:
(ulong) trx->mysql_n_tables_locked);
}
newline = TRUE;
bool newline = true;
if (trx->in_rollback) { /* dirty read for performance reasons */
fputs("ROLLING BACK ", f);
} else if (trx->lock.wait_lock) {
fputs("LOCK WAIT ", f);
} else {
newline = FALSE;
newline = false;
}
if (n_trx_locks > 0 || heap_size > 400) {
newline = TRUE;
newline = true;
fprintf(f, "%lu lock struct(s), heap size %lu,"
" %lu row lock(s)",
@ -1771,7 +1763,7 @@ state_ok:
}
if (trx->undo_no != 0) {
newline = TRUE;
newline = true;
fprintf(f, ", undo log entries " TRX_ID_FMT, trx->undo_no);
}
@ -2150,7 +2142,7 @@ trx_start_internal_low(
/* Ensure it is not flagged as an auto-commit-non-locking
transaction. */
trx->will_lock = 1;
trx->will_lock = true;
trx->internal = true;
@ -2166,7 +2158,7 @@ trx_start_internal_read_only_low(
/* Ensure it is not flagged as an auto-commit-non-locking
transaction. */
trx->will_lock = 1;
trx->will_lock = true;
trx->internal = true;
@ -2181,8 +2173,6 @@ void trx_start_for_ddl_low(trx_t *trx)
/* Flag this transaction as a dictionary operation, so that
the data dictionary will be locked in crash recovery. */
trx->dict_operation= true;
/* Ensure it is not flagged as an auto-commit-non-locking transaction. */
trx->will_lock= 1;
trx_start_internal_low(trx);
}
@ -2199,7 +2189,7 @@ trx_set_rw_mode(
trx_t* trx) /*!< in/out: transaction that is RW */
{
ut_ad(trx->rsegs.m_redo.rseg == 0);
ut_ad(!trx_is_autocommit_non_locking(trx));
ut_ad(!trx->is_autocommit_non_locking());
ut_ad(!trx->read_only);
ut_ad(trx->id == 0);