2010-07-02 18:07:57 +02:00
set global storage_engine=myisam;
set session storage_engine=myisam;
2001-10-03 23:19:58 +02:00
drop table if exists t1,t2,t3,t4,t5,t6;
2003-06-27 15:29:10 +02:00
drop database if exists mysqltest;
2001-09-28 07:05:54 +02:00
create table t1 (a int not null primary key auto_increment, message char(20));
create table t2 (a int not null primary key auto_increment, message char(20));
INSERT INTO t1 (message) VALUES ("Testing"),("table"),("t1");
INSERT INTO t2 (message) VALUES ("Testing"),("table"),("t2");
2003-12-10 05:31:42 +01:00
create table t3 (a int not null, b char(20), key(a)) engine=MERGE UNION=(t1,t2);
2001-09-28 07:05:54 +02:00
select * from t3;
2000-12-28 02:56:38 +01:00
a b
1 Testing
2 table
3 t1
1 Testing
2 table
3 t2
2001-09-28 07:05:54 +02:00
select * from t3 order by a desc;
2000-12-28 02:56:38 +01:00
a b
3 t1
3 t2
2 table
2 table
1 Testing
1 Testing
2001-09-28 07:05:54 +02:00
drop table t3;
insert into t1 select NULL,message from t2;
insert into t2 select NULL,message from t1;
insert into t1 select NULL,message from t2;
insert into t2 select NULL,message from t1;
insert into t1 select NULL,message from t2;
insert into t2 select NULL,message from t1;
insert into t1 select NULL,message from t2;
insert into t2 select NULL,message from t1;
insert into t1 select NULL,message from t2;
insert into t2 select NULL,message from t1;
insert into t1 select NULL,message from t2;
2003-12-10 05:31:42 +01:00
create table t3 (a int not null, b char(20), key(a)) engine=MERGE UNION=(test.t1,test.t2);
2001-09-28 07:05:54 +02:00
explain select * from t3 where a < 10;
2002-09-26 22:08:22 +02:00
id select_type table type possible_keys key key_len ref rows Extra
2002-11-21 21:25:53 +01:00
1 SIMPLE t3 range a a 4 NULL 18 Using where
2001-09-28 07:05:54 +02:00
explain select * from t3 where a > 10 and a < 20;
2002-09-26 22:08:22 +02:00
id select_type table type possible_keys key key_len ref rows Extra
2005-05-13 11:08:08 +02:00
1 SIMPLE t3 range a a 4 NULL 17 Using where
2001-09-28 07:05:54 +02:00
select * from t3 where a = 10;
2000-12-28 02:56:38 +01:00
a b
10 Testing
10 Testing
2001-09-28 07:05:54 +02:00
select * from t3 where a < 10;
2000-12-28 02:56:38 +01:00
a b
1 Testing
1 Testing
2 table
2 table
3 t1
3 t2
4 Testing
4 Testing
5 table
5 table
6 t2
2006-02-11 19:51:43 +01:00
6 t1
2000-12-28 02:56:38 +01:00
7 Testing
7 Testing
8 table
8 table
9 t2
9 t2
2001-09-28 07:05:54 +02:00
select * from t3 where a > 10 and a < 20;
2000-12-28 02:56:38 +01:00
a b
11 table
11 table
12 t1
12 t1
13 Testing
13 Testing
14 table
14 table
15 t2
15 t2
16 Testing
16 Testing
17 table
17 table
18 t2
18 t2
19 Testing
19 Testing
2001-09-28 07:05:54 +02:00
explain select a from t3 order by a desc limit 10;
2002-09-26 22:08:22 +02:00
id select_type table type possible_keys key key_len ref rows Extra
2007-08-02 21:45:56 +02:00
1 SIMPLE t3 index NULL a 4 NULL 10 Using index
2001-09-28 07:05:54 +02:00
select a from t3 order by a desc limit 10;
2000-12-28 02:56:38 +01:00
a
699
698
697
696
695
694
693
692
691
690
2001-09-28 07:05:54 +02:00
select a from t3 order by a desc limit 300,10;
2000-12-28 02:56:38 +01:00
a
416
415
415
414
414
413
413
412
412
411
2001-10-03 23:19:58 +02:00
delete from t3 where a=3;
select * from t3 where a < 10;
a b
1 Testing
1 Testing
2 table
2 table
4 Testing
4 Testing
5 table
5 table
6 t2
6 t1
7 Testing
7 Testing
8 table
8 table
9 t2
9 t2
delete from t3 where a >= 6 and a <= 8;
select * from t3 where a < 10;
a b
1 Testing
1 Testing
2 table
2 table
4 Testing
4 Testing
5 table
5 table
9 t2
9 t2
update t3 set a=3 where a=9;
select * from t3 where a < 10;
a b
1 Testing
1 Testing
2 table
2 table
3 t2
3 t2
4 Testing
4 Testing
5 table
5 table
update t3 set a=6 where a=7;
select * from t3 where a < 10;
a b
1 Testing
1 Testing
2 table
2 table
3 t2
3 t2
4 Testing
4 Testing
5 table
5 table
2001-09-28 07:05:54 +02:00
show create table t3;
2001-02-17 18:04:33 +01:00
Table Create Table
t3 CREATE TABLE `t3` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
2006-02-22 10:09:59 +01:00
`b` char(20) DEFAULT NULL,
2001-02-17 18:04:33 +01:00
KEY `a` (`a`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`t2`)
create table t4 (a int not null, b char(10), key(a)) engine=MERGE UNION=(t1,t2);
2001-09-28 07:05:54 +02:00
select * from t4;
2006-10-06 07:54:47 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2003-06-27 15:29:10 +02:00
alter table t4 add column c int;
2006-10-06 07:54:47 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
Table definition cache, part 2
The table opening process now works the following way:
- Create common TABLE_SHARE object
- Read the .frm file and unpack it into the TABLE_SHARE object
- Create a TABLE object based on the information in the TABLE_SHARE
object and open a handler to the table object
Other noteworthy changes:
- In TABLE_SHARE the most common strings are now LEX_STRING's
- Better error message when table is not found
- Variable table_cache is now renamed 'table_open_cache'
- New variable 'table_definition_cache' that is the number of table defintions that will be cached
- strxnmov() calls are now fixed to avoid overflows
- strxnmov() will now always add one end \0 to result
- engine objects are now created with a TABLE_SHARE object instead of a TABLE object.
- After creating a field object one must call field->init(table) before using it
- For a busy system this change will give you:
- Less memory usage for table object
- Faster opening of tables (if it's has been in use or is in table definition cache)
- Allow you to cache many table definitions objects
- Faster drop of table
mysql-test/mysql-test-run.sh:
Fixed some problems with --gdb option
Test both with socket and tcp/ip port that all old servers are killed
mysql-test/r/flush_table.result:
More tests with lock table with 2 threads + flush table
mysql-test/r/information_schema.result:
Removed old (now wrong) result
mysql-test/r/innodb.result:
Better error messages (thanks to TDC patch)
mysql-test/r/merge.result:
Extra flush table test
mysql-test/r/ndb_bitfield.result:
Better error messages (thanks to TDC patch)
mysql-test/r/ndb_partition_error.result:
Better error messages (thanks to TDC patch)
mysql-test/r/query_cache.result:
Remove tables left from old tests
mysql-test/r/temp_table.result:
Test truncate with temporary tables
mysql-test/r/variables.result:
Table_cache -> Table_open_cache
mysql-test/t/flush_table.test:
More tests with lock table with 2 threads + flush table
mysql-test/t/merge.test:
Extra flush table test
mysql-test/t/multi_update.test:
Added 'sleep' to make test predictable
mysql-test/t/query_cache.test:
Remove tables left from old tests
mysql-test/t/temp_table.test:
Test truncate with temporary tables
mysql-test/t/variables.test:
Table_cache -> Table_open_cache
mysql-test/valgrind.supp:
Remove warning that may happens becasue threads dies in different order
mysys/hash.c:
Fixed wrong DBUG_PRINT
mysys/mf_dirname.c:
More DBUG
mysys/mf_pack.c:
Better comment
mysys/mf_tempdir.c:
More DBUG
Ensure that we call cleanup_dirname() on all temporary directory paths.
If we don't do this, we will get a failure when comparing temporary table
names as in some cases the temporary table name is run through convert_dirname())
mysys/my_alloc.c:
Indentation fix
sql/examples/ha_example.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/examples/ha_example.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/examples/ha_tina.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/examples/ha_tina.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/field.cc:
Update for table definition cache:
- Field creation now takes TABLE_SHARE instead of TABLE as argument
(This is becasue field definitions are now cached in TABLE_SHARE)
When a field is created, one now must call field->init(TABLE) before using it
- Use s->db instead of s->table_cache_key
- Added Field::clone() to create a field in TABLE from a field in TABLE_SHARE
- make_field() takes TABLE_SHARE as argument instead of TABLE
- move_field() -> move_field_offset()
sql/field.h:
Update for table definition cache:
- Field creation now takes TABLE_SHARE instead of TABLE as argument
(This is becasue field definitions are now cached in TABLE_SHARE)
When a field is created, one now must call field->init(TABLE) before using it
- Added Field::clone() to create a field in TABLE from a field in TABLE_SHARE
- make_field() takes TABLE_SHARE as argument instead of TABLE
- move_field() -> move_field_offset()
sql/ha_archive.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_archive.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_berkeley.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Changed name of argument create() to not hide internal 'table' variable.
table->s -> table_share
sql/ha_berkeley.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_blackhole.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_blackhole.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_federated.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Fixed comments
Remove index variable and replace with pointers (simple optimization)
move_field() -> move_field_offset()
Removed some strlen() calls
sql/ha_federated.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_heap.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Simplify delete_table() and create() as the given file names are now without extension
sql/ha_heap.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_innodb.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_innodb.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_myisam.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Remove not needed fn_format()
Fixed for new table->s structure
sql/ha_myisam.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_myisammrg.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Don't set 'is_view' for MERGE tables
Use new interface to find_temporary_table()
sql/ha_myisammrg.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Added flag HA_NO_COPY_ON_ALTER
sql/ha_ndbcluster.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Fixed wrong calls to strxnmov()
Give error HA_ERR_TABLE_DEF_CHANGED if table definition has changed
drop_table -> intern_drop_table()
table->s -> table_share
Move part_info to TABLE
Fixed comments & DBUG print's
New arguments to print_error()
sql/ha_ndbcluster.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
sql/ha_partition.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
We can't set up or use part_info when creating handler as there is not yet any table object
New ha_intialise() to work with TDC (Done by Mikael)
sql/ha_partition.h:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
Got set_part_info() from Mikael
sql/handler.cc:
We new use TABLE_SHARE instead of TABLE when creating engine handlers
ha_delete_table() now also takes database as an argument
handler::ha_open() now takes TABLE as argument
ha_open() now calls ha_allocate_read_write_set()
Simplify ha_allocate_read_write_set()
Remove ha_deallocate_read_write_set()
Use table_share (Cached by table definition cache)
sql/handler.h:
New table flag: HA_NO_COPY_ON_ALTER (used by merge tables)
Remove ha_deallocate_read_write_set()
get_new_handler() now takes TABLE_SHARE as argument
ha_delete_table() now gets database as argument
sql/item.cc:
table_name and db are now LEX_STRING objects
When creating fields, we have now have to call field->init(table)
move_field -> move_field_offset()
sql/item.h:
tmp_table_field_from_field_type() now takes an extra paramenter 'fixed_length' to allow one to force usage of CHAR
instead of BLOB
sql/item_cmpfunc.cc:
Fixed call to tmp_table_field_from_field_type()
sql/item_create.cc:
Assert if new not handled cast type
sql/item_func.cc:
When creating fields, we have now have to call field->init(table)
dummy_table used by 'sp' now needs a TABLE_SHARE object
sql/item_subselect.cc:
Trivial code cleanups
sql/item_sum.cc:
When creating fields, we have now have to call field->init(table)
sql/item_timefunc.cc:
Item_func_str_to_date::tmp_table_field() now replaced by call to
tmp_table_field_from_field_type() (see item_timefunc.h)
sql/item_timefunc.h:
Simply tmp_table_field()
sql/item_uniq.cc:
When creating fields, we have now have to call field->init(table)
sql/key.cc:
Added 'KEY' argument to 'find_ref_key' to simplify code
sql/lock.cc:
More debugging
Use create_table_def_key() to create key for table cache
Allocate TABLE_SHARE properly when creating name lock
Fix that locked_table_name doesn't test same table twice
sql/mysql_priv.h:
New functions for table definition cache
New interfaces to a lot of functions.
New faster interface to find_temporary_table() and close_temporary_table()
sql/mysqld.cc:
Added support for table definition cache of size 'table_def_size'
Fixed som calls to strnmov()
Changed name of 'table_cache' to 'table_open_cache'
sql/opt_range.cc:
Use new interfaces
Fixed warnings from valgrind
sql/parse_file.cc:
Safer calls to strxnmov()
Fixed typo
sql/set_var.cc:
Added variable 'table_definition_cache'
Variable table_cache renamed to 'table_open_cache'
sql/slave.cc:
Use new interface
sql/sp.cc:
Proper use of TABLE_SHARE
sql/sp_head.cc:
Remove compiler warnings
We have now to call field->init(table)
sql/sp_head.h:
Pointers to parsed strings are now const
sql/sql_acl.cc:
table_name is now a LEX_STRING
sql/sql_base.cc:
Main implementation of table definition cache
(The #ifdef's are there for the future when table definition cache will replace open table cache)
Now table definitions are cached indepndent of open tables, which will speed up things when a table is in use at once from several places
Views are not yet cached; For the moment we only cache if a table is a view or not.
Faster implementation of find_temorary_table()
Replace 'wait_for_refresh()' with the more general function 'wait_for_condition()'
Drop table is slightly faster as we can use the table definition cache to know the type of the table
sql/sql_cache.cc:
table_cache_key and table_name are now LEX_STRING
'sDBUG print fixes
sql/sql_class.cc:
table_cache_key is now a LEX_STRING
safer strxnmov()
sql/sql_class.h:
Added number of open table shares (table definitions)
sql/sql_db.cc:
safer strxnmov()
sql/sql_delete.cc:
Use new interface to find_temporary_table()
sql/sql_derived.cc:
table_name is now a LEX_STRING
sql/sql_handler.cc:
TABLE_SHARE->db and TABLE_SHARE->table_name are now LEX_STRING's
sql/sql_insert.cc:
TABLE_SHARE->db and TABLE_SHARE->table_name are now LEX_STRING's
sql/sql_lex.cc:
Make parsed string a const (to quickly find out if anything is trying to change the query string)
sql/sql_lex.h:
Make parsed string a const (to quickly find out if anything is trying to change the query string)
sql/sql_load.cc:
Safer strxnmov()
sql/sql_parse.cc:
Better error if wrong DB name
sql/sql_partition.cc:
part_info moved to TABLE from TABLE_SHARE
Indentation changes
sql/sql_select.cc:
Indentation fixes
Call field->init(TABLE) for new created fields
Update create_tmp_table() to use TABLE_SHARE properly
sql/sql_select.h:
Call field->init(TABLE) for new created fields
sql/sql_show.cc:
table_name is now a LEX_STRING
part_info moved to TABLE
sql/sql_table.cc:
Use table definition cache to speed up delete of tables
Fixed calls to functions with new interfaces
Don't use 'share_not_to_be_used'
Instead of doing openfrm() when doing repair, we now have to call
get_table_share() followed by open_table_from_share().
Replace some fn_format() with faster unpack_filename().
Safer strxnmov()
part_info is now in TABLE
Added Mikaels patch for partition and ALTER TABLE
Instead of using 'TABLE_SHARE->is_view' use 'table_flags() & HA_NO_COPY_ON_ALTER
sql/sql_test.cc:
table_name and table_cache_key are now LEX_STRING's
sql/sql_trigger.cc:
TABLE_SHARE->db and TABLE_SHARE->table_name are now LEX_STRING's
safer strxnmov()
Removed compiler warnings
sql/sql_update.cc:
Call field->init(TABLE) after field is created
sql/sql_view.cc:
safer strxnmov()
Create common TABLE_SHARE object for views to allow us to cache if table is a view
sql/structs.h:
Added SHOW_TABLE_DEFINITIONS
sql/table.cc:
Creation and destruct of TABLE_SHARE objects that are common for many TABLE objects
The table opening process now works the following way:
- Create common TABLE_SHARE object
- Read the .frm file and unpack it into the TABLE_SHARE object
- Create a TABLE object based on the information in the TABLE_SHARE
object and open a handler to the table object
open_table_def() is written in such a way that it should be trival to add parsing of the .frm files in new formats
sql/table.h:
TABLE objects for the same database table now share a common TABLE_SHARE object
In TABLE_SHARE the most common strings are now LEX_STRING's
sql/unireg.cc:
Changed arguments to rea_create_table() to have same order as other functions
Call field->init(table) for new created fields
sql/unireg.h:
Added OPEN_VIEW
strings/strxnmov.c:
Change strxnmov() to always add end \0
This makes usage of strxnmov() safer as most of MySQL code assumes that strxnmov() will create a null terminated string
2005-11-23 21:45:02 +01:00
flush tables;
select * from t4;
2006-10-06 08:09:58 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2003-06-27 15:29:10 +02:00
create database mysqltest;
create table mysqltest.t6 (a int not null primary key auto_increment, message char(20));
2003-12-10 05:31:42 +01:00
create table t5 (a int not null, b char(20), key(a)) engine=MERGE UNION=(test.t1,mysqltest.t6);
2003-06-27 15:29:10 +02:00
show create table t5;
Table Create Table
t5 CREATE TABLE `t5` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
2006-02-22 10:09:59 +01:00
`b` char(20) DEFAULT NULL,
2003-06-27 15:29:10 +02:00
KEY `a` (`a`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`mysqltest`.`t6`)
alter table t5 engine=myisam;
2003-06-27 15:29:10 +02:00
drop table t5, mysqltest.t6;
drop database mysqltest;
2003-07-01 14:49:32 +02:00
drop table t4,t3,t1,t2;
2003-12-10 05:31:42 +01:00
create table t1 (c char(10)) engine=myisam;
create table t2 (c char(10)) engine=myisam;
create table t3 (c char(10)) union=(t1,t2) engine=merge;
2001-09-28 07:05:54 +02:00
insert into t1 (c) values ('test1');
insert into t1 (c) values ('test1');
insert into t1 (c) values ('test1');
insert into t2 (c) values ('test2');
insert into t2 (c) values ('test2');
insert into t2 (c) values ('test2');
select * from t3;
2000-12-28 02:56:38 +01:00
c
test1
test1
test1
test2
test2
test2
2001-09-28 07:05:54 +02:00
select * from t3;
2000-12-28 02:56:38 +01:00
c
test1
test1
test1
test2
test2
test2
2001-09-28 07:05:54 +02:00
delete from t3 where 1=1;
select * from t3;
2000-12-28 02:56:38 +01:00
c
2001-09-28 07:05:54 +02:00
select * from t1;
2000-12-28 02:56:38 +01:00
c
2001-09-28 07:05:54 +02:00
drop table t3,t2,t1;
CREATE TABLE t1 (incr int not null, othr int not null, primary key(incr));
CREATE TABLE t2 (incr int not null, othr int not null, primary key(incr));
CREATE TABLE t3 (incr int not null, othr int not null, primary key(incr))
2003-12-10 05:31:42 +01:00
ENGINE=MERGE UNION=(t1,t2);
2001-09-28 07:05:54 +02:00
SELECT * from t3;
2000-12-28 02:56:38 +01:00
incr othr
2001-09-28 07:05:54 +02:00
INSERT INTO t1 VALUES ( 1,10),( 3,53),( 5,21),( 7,12),( 9,17);
INSERT INTO t2 VALUES ( 2,24),( 4,33),( 6,41),( 8,26),( 0,32);
INSERT INTO t1 VALUES (11,20),(13,43),(15,11),(17,22),(19,37);
INSERT INTO t2 VALUES (12,25),(14,31),(16,42),(18,27),(10,30);
SELECT * from t3 where incr in (1,2,3,4) order by othr;
2000-12-28 02:56:38 +01:00
incr othr
1 10
2 24
4 33
3 53
2001-09-28 07:05:54 +02:00
alter table t3 UNION=(t1);
select count(*) from t3;
2001-01-28 20:35:50 +01:00
count(*)
10
2001-09-28 07:05:54 +02:00
alter table t3 UNION=(t1,t2);
select count(*) from t3;
2001-01-28 20:35:50 +01:00
count(*)
20
2003-12-10 05:31:42 +01:00
alter table t3 ENGINE=MYISAM;
2001-09-28 07:05:54 +02:00
select count(*) from t3;
2001-01-28 20:35:50 +01:00
count(*)
20
2001-09-28 07:05:54 +02:00
drop table t3;
CREATE TABLE t3 (incr int not null, othr int not null, primary key(incr))
2003-12-10 05:31:42 +01:00
ENGINE=MERGE UNION=(t1,t2);
2001-09-28 07:05:54 +02:00
show create table t3;
2001-03-06 14:24:08 +01:00
Table Create Table
t3 CREATE TABLE `t3` (
2004-10-02 21:20:08 +02:00
`incr` int(11) NOT NULL,
`othr` int(11) NOT NULL,
2006-02-02 20:39:28 +01:00
PRIMARY KEY (`incr`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`t2`)
2001-09-28 07:05:54 +02:00
alter table t3 drop primary key;
show create table t3;
2001-03-06 14:24:08 +01:00
Table Create Table
t3 CREATE TABLE `t3` (
2004-10-02 21:20:08 +02:00
`incr` int(11) NOT NULL,
`othr` int(11) NOT NULL
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`t2`)
2001-09-28 07:05:54 +02:00
drop table t3,t2,t1;
2003-12-10 05:31:42 +01:00
create table t1 (a int not null, key(a)) engine=merge;
2001-09-28 07:05:54 +02:00
select * from t1;
2010-05-31 15:38:02 +02:00
a
2001-09-28 07:05:54 +02:00
drop table t1;
create table t1 (a int not null, b int not null, key(a,b));
create table t2 (a int not null, b int not null, key(a,b));
2003-12-10 05:31:42 +01:00
create table t3 (a int not null, b int not null, key(a,b)) ENGINE=MERGE UNION=(t1,t2);
2001-09-28 07:05:54 +02:00
insert into t1 values (1,2),(2,1),(0,0),(4,4),(5,5),(6,6);
insert into t2 values (1,1),(2,2),(0,0),(4,4),(5,5),(6,6);
flush tables;
select * from t3 where a=1 order by b limit 2;
2001-05-16 23:06:30 +02:00
a b
1 1
1 2
2001-09-28 07:05:54 +02:00
drop table t3,t1,t2;
2002-07-21 23:55:32 +02:00
create table t1 (a int not null, b int not null auto_increment, primary key(a,b));
create table t2 (a int not null, b int not null auto_increment, primary key(a,b));
2001-10-03 23:19:58 +02:00
create table t3 (a int not null, b int not null, key(a,b)) UNION=(t1,t2) INSERT_METHOD=NO;
2003-12-10 05:31:42 +01:00
create table t4 (a int not null, b int not null, key(a,b)) ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=NO;
create table t5 (a int not null, b int not null auto_increment, primary key(a,b)) ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=FIRST;
create table t6 (a int not null, b int not null auto_increment, primary key(a,b)) ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=LAST;
2001-10-03 23:19:58 +02:00
show create table t3;
Table Create Table
t3 CREATE TABLE `t3` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
`b` int(11) NOT NULL,
2001-10-03 23:19:58 +02:00
KEY `a` (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MyISAM DEFAULT CHARSET=latin1
2001-10-03 23:19:58 +02:00
show create table t4;
Table Create Table
t4 CREATE TABLE `t4` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
`b` int(11) NOT NULL,
2001-10-03 23:19:58 +02:00
KEY `a` (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`t2`)
2001-10-03 23:19:58 +02:00
show create table t5;
Table Create Table
t5 CREATE TABLE `t5` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
2006-02-22 10:09:59 +01:00
`b` int(11) NOT NULL AUTO_INCREMENT,
2006-02-02 20:39:28 +01:00
PRIMARY KEY (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=FIRST UNION=(`t1`,`t2`)
2001-10-03 23:19:58 +02:00
show create table t6;
Table Create Table
t6 CREATE TABLE `t6` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
2006-02-22 10:09:59 +01:00
`b` int(11) NOT NULL AUTO_INCREMENT,
2006-02-02 20:39:28 +01:00
PRIMARY KEY (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
2002-07-21 23:55:32 +02:00
insert into t1 values (1,NULL),(1,NULL),(1,NULL),(1,NULL);
insert into t2 values (2,NULL),(2,NULL),(2,NULL),(2,NULL);
2001-10-03 23:19:58 +02:00
select * from t3 order by b,a limit 3;
a b
select * from t4 order by b,a limit 3;
a b
1 1
2 1
1 2
select * from t5 order by b,a limit 3,3;
a b
2 2
1 3
2 3
select * from t6 order by b,a limit 6,3;
a b
1 4
2 4
insert into t5 values (5,1),(5,2);
insert into t6 values (6,1),(6,2);
select * from t1 order by a,b;
a b
1 1
1 2
1 3
1 4
5 1
5 2
select * from t2 order by a,b;
a b
2 1
2 2
2 3
2 4
6 1
6 2
select * from t4 order by a,b;
a b
1 1
1 2
1 3
1 4
2 1
2 2
2 3
2 4
5 1
5 2
6 1
6 2
insert into t3 values (3,1),(3,2),(3,3),(3,4);
select * from t3 order by a,b;
a b
3 1
3 2
3 3
3 4
alter table t4 UNION=(t1,t2,t3);
show create table t4;
Table Create Table
t4 CREATE TABLE `t4` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
`b` int(11) NOT NULL,
2001-10-03 23:19:58 +02:00
KEY `a` (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t1`,`t2`,`t3`)
2001-10-03 23:19:58 +02:00
select * from t4 order by a,b;
a b
1 1
1 2
1 3
1 4
2 1
2 2
2 3
2 4
3 1
3 2
3 3
3 4
5 1
5 2
6 1
6 2
alter table t4 INSERT_METHOD=FIRST;
show create table t4;
Table Create Table
t4 CREATE TABLE `t4` (
2004-10-02 21:20:08 +02:00
`a` int(11) NOT NULL,
`b` int(11) NOT NULL,
2001-10-03 23:19:58 +02:00
KEY `a` (`a`,`b`)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=FIRST UNION=(`t1`,`t2`,`t3`)
2001-10-03 23:19:58 +02:00
insert into t4 values (4,1),(4,2);
select * from t1 order by a,b;
a b
1 1
1 2
1 3
1 4
4 1
4 2
5 1
5 2
select * from t2 order by a,b;
a b
2 1
2 2
2 3
2 4
6 1
6 2
select * from t3 order by a,b;
a b
3 1
3 2
3 3
3 4
select * from t4 order by a,b;
a b
1 1
1 2
1 3
1 4
2 1
2 2
2 3
2 4
3 1
3 2
3 3
3 4
4 1
4 2
5 1
5 2
6 1
6 2
select * from t5 order by a,b;
a b
1 1
1 2
1 3
1 4
2 1
2 2
2 3
2 4
4 1
4 2
5 1
5 2
6 1
6 2
2002-07-21 23:55:32 +02:00
select 1;
1
1
insert into t5 values (1,NULL),(5,NULL);
insert into t6 values (2,NULL),(6,NULL);
select * from t1 order by a,b;
a b
1 1
1 2
1 3
1 4
1 5
4 1
4 2
5 1
5 2
5 3
select * from t2 order by a,b;
a b
2 1
2 2
2 3
2 4
2 5
6 1
6 2
6 3
select * from t5 order by a,b;
a b
1 1
1 2
1 3
1 4
1 5
2 1
2 2
2 3
2 4
2 5
4 1
4 2
5 1
5 2
5 3
6 1
6 2
6 3
select * from t6 order by a,b;
a b
1 1
1 2
1 3
1 4
1 5
2 1
2 2
2 3
2 4
2 5
4 1
4 2
5 1
5 2
5 3
6 1
6 2
6 3
2003-07-01 14:49:32 +02:00
insert into t1 values (99,NULL);
select * from t4 where a+0 > 90;
a b
99 1
2004-06-25 15:12:13 +02:00
insert t5 values (1,1);
2006-01-23 12:17:05 +01:00
ERROR 23000: Duplicate entry '1-1' for key 'PRIMARY'
2004-06-25 15:12:13 +02:00
insert t6 values (2,1);
2006-01-23 12:17:05 +01:00
ERROR 23000: Duplicate entry '2-1' for key 'PRIMARY'
2004-06-25 15:12:13 +02:00
insert t5 values (1,1) on duplicate key update b=b+10;
insert t6 values (2,1) on duplicate key update b=b+20;
select * from t5 where a < 3;
a b
1 2
1 3
1 4
1 5
1 11
2 2
2 3
2 4
2 5
2 21
2003-07-01 14:49:32 +02:00
drop table t6, t5, t4, t3, t2, t1;
2003-12-10 05:31:42 +01:00
CREATE TABLE t1 ( a int(11) NOT NULL default '0', b int(11) NOT NULL default '0', PRIMARY KEY (a,b)) ENGINE=MyISAM;
2002-06-15 19:59:42 +02:00
INSERT INTO t1 VALUES (1,1), (2,1);
2003-12-10 05:31:42 +01:00
CREATE TABLE t2 ( a int(11) NOT NULL default '0', b int(11) NOT NULL default '0', PRIMARY KEY (a,b)) ENGINE=MyISAM;
2002-06-15 19:59:42 +02:00
INSERT INTO t2 VALUES (1,2), (2,2);
2003-12-10 05:31:42 +01:00
CREATE TABLE t3 ( a int(11) NOT NULL default '0', b int(11) NOT NULL default '0', KEY a (a,b)) ENGINE=MRG_MyISAM UNION=(t1,t2);
2003-01-06 00:48:59 +01:00
select max(b) from t3 where a = 2;
2002-06-15 19:59:42 +02:00
max(b)
2003-01-17 12:48:23 +01:00
2
2002-06-18 16:38:06 +02:00
select max(b) from t1 where a = 2;
max(b)
1
2003-07-01 14:49:32 +02:00
drop table t3,t1,t2;
2002-11-05 18:38:27 +01:00
create table t1 (a int not null);
create table t2 (a int not null);
insert into t1 values (1);
insert into t2 values (2);
2003-12-10 05:31:42 +01:00
create temporary table t3 (a int not null) ENGINE=MERGE UNION=(t1,t2);
2002-11-05 18:38:27 +01:00
select * from t3;
2013-03-27 10:03:28 +01:00
a
2010-07-02 18:07:57 +02:00
1
2
2002-11-05 18:38:27 +01:00
create temporary table t4 (a int not null);
create temporary table t5 (a int not null);
insert into t4 values (1);
insert into t5 values (2);
2003-12-10 05:31:42 +01:00
create temporary table t6 (a int not null) ENGINE=MERGE UNION=(t4,t5);
2002-11-05 18:38:27 +01:00
select * from t6;
a
1
2
2003-07-01 14:49:32 +02:00
drop table t6, t3, t1, t2, t4, t5;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
create temporary table t1 (a int not null);
create temporary table t2 (a int not null);
insert into t1 values (1);
insert into t2 values (2);
create table t3 (a int not null) ENGINE=MERGE UNION=(t1,t2);
select * from t3;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
drop table t3, t2, t1;
create table t1 (a int not null);
create temporary table t2 (a int not null);
insert into t1 values (1);
insert into t2 values (2);
create table t3 (a int not null) ENGINE=MERGE UNION=(t1,t2);
select * from t3;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
drop table t3;
create temporary table t3 (a int not null) ENGINE=MERGE UNION=(t1,t2);
select * from t3;
2010-07-02 18:07:57 +02:00
a
1
2
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
drop table t3, t2, t1;
# CREATE...SELECT is not implemented for MERGE tables.
CREATE TEMPORARY TABLE t1 (c1 INT NOT NULL);
CREATE TEMPORARY TABLE t2 (c1 INT NOT NULL);
CREATE TABLE t3 (c1 INT NOT NULL);
INSERT INTO t3 VALUES (3), (33);
LOCK TABLES t3 READ;
CREATE TEMPORARY TABLE t4 (c1 INT NOT NULL) ENGINE=MERGE UNION=(t1,t2)
INSERT_METHOD=LAST SELECT * FROM t3;
ERROR HY000: 'test.t4' is not BASE TABLE
SELECT * FROM t4;
ERROR HY000: Table 't4' was not locked with LOCK TABLES
UNLOCK TABLES;
CREATE TEMPORARY TABLE t4 (c1 INT NOT NULL) ENGINE=MERGE UNION=(t1,t2)
INSERT_METHOD=LAST;
INSERT INTO t4 SELECT * FROM t3;
# Alter temporary MERGE table.
ALTER TABLE t4 UNION=(t1);
LOCK TABLES t4 WRITE;
# Alter temporary MERGE table under LOCk tables.
ALTER TABLE t4 UNION=(t1,t2);
UNLOCK TABLES;
# MERGE table and function.
CREATE FUNCTION f1 () RETURNS INT RETURN (SELECT max(c1) FROM t3);
SELECT * FROM t4 WHERE c1 < f1();
c1
3
DROP FUNCTION f1;
DROP TABLE t4, t3, t2, t1;
2002-11-17 18:58:03 +01:00
CREATE TABLE t1 (
fileset_id tinyint(3) unsigned NOT NULL default '0',
file_code varchar(32) NOT NULL default '',
fileset_root_id tinyint(3) unsigned NOT NULL default '0',
PRIMARY KEY (fileset_id,file_code),
KEY files (fileset_id,fileset_root_id)
2003-12-10 05:31:42 +01:00
) ENGINE=MyISAM;
2002-11-17 18:58:03 +01:00
INSERT INTO t1 VALUES (2, '0000000111', 1), (2, '0000000112', 1), (2, '0000000113', 1),
(2, '0000000114', 1), (2, '0000000115', 1), (2, '0000000116', 1), (2, '0000000117', 1),
(2, '0000000118', 1), (2, '0000000119', 1), (2, '0000000120', 1);
CREATE TABLE t2 (
fileset_id tinyint(3) unsigned NOT NULL default '0',
file_code varchar(32) NOT NULL default '',
fileset_root_id tinyint(3) unsigned NOT NULL default '0',
PRIMARY KEY (fileset_id,file_code),
KEY files (fileset_id,fileset_root_id)
2003-12-10 05:31:42 +01:00
) ENGINE=MRG_MyISAM UNION=(t1);
2002-11-17 18:58:03 +01:00
EXPLAIN SELECT * FROM t2 IGNORE INDEX (files) WHERE fileset_id = 2
AND file_code BETWEEN '0000000115' AND '0000000120' LIMIT 1;
2002-11-21 21:25:53 +01:00
id select_type table type possible_keys key key_len ref rows Extra
2004-12-06 01:00:37 +01:00
1 SIMPLE t2 range PRIMARY PRIMARY 35 NULL 5 Using where
2002-11-17 18:58:03 +01:00
EXPLAIN SELECT * FROM t2 WHERE fileset_id = 2
AND file_code BETWEEN '0000000115' AND '0000000120' LIMIT 1;
2002-11-21 21:25:53 +01:00
id select_type table type possible_keys key key_len ref rows Extra
2004-12-06 01:00:37 +01:00
1 SIMPLE t2 range PRIMARY,files PRIMARY 35 NULL 5 Using where
2002-11-17 18:58:03 +01:00
EXPLAIN SELECT * FROM t1 WHERE fileset_id = 2
AND file_code BETWEEN '0000000115' AND '0000000120' LIMIT 1;
2002-11-21 21:25:53 +01:00
id select_type table type possible_keys key key_len ref rows Extra
2011-11-21 14:16:16 +01:00
1 SIMPLE t1 range PRIMARY,files PRIMARY 35 NULL 5 Using index condition
2002-11-17 18:58:03 +01:00
EXPLAIN SELECT * FROM t2 WHERE fileset_id = 2
AND file_code = '0000000115' LIMIT 1;
2002-11-21 21:25:53 +01:00
id select_type table type possible_keys key key_len ref rows Extra
2006-10-16 17:09:58 +02:00
1 SIMPLE t2 const PRIMARY,files PRIMARY 35 const,const 1
2003-01-06 00:48:59 +01:00
DROP TABLE t2, t1;
2003-07-03 10:55:36 +02:00
create table t1 (x int, y int, index xy(x, y));
create table t2 (x int, y int, index xy(x, y));
2003-12-10 05:31:42 +01:00
create table t3 (x int, y int, index xy(x, y)) engine=merge union=(t1,t2);
2003-07-03 10:55:36 +02:00
insert into t1 values(1, 2);
insert into t2 values(1, 3);
select * from t3 where x = 1 and y < 5 order by y;
x y
1 2
1 3
select * from t3 where x = 1 and y < 5 order by y desc;
x y
2003-07-04 14:42:57 +02:00
1 3
1 2
2003-07-03 10:55:36 +02:00
drop table t1,t2,t3;
2004-08-27 13:43:01 +02:00
create table t1 (a int);
create table t2 (a int);
insert into t1 values (0);
insert into t2 values (1);
create table t3 engine=merge union=(t1, t2) select * from t1;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
ERROR HY000: 'test.t3' is not BASE TABLE
2004-08-27 13:43:01 +02:00
create table t3 engine=merge union=(t1, t2) select * from t2;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
ERROR HY000: 'test.t3' is not BASE TABLE
2004-09-10 12:09:27 +02:00
create table t3 engine=merge union=(t1, t2) select (select max(a) from t2);
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
ERROR HY000: 'test.t3' is not BASE TABLE
2004-08-27 13:43:01 +02:00
drop table t1, t2;
2005-09-23 10:15:11 +02:00
create table t1 (
2005-09-27 15:26:16 +02:00
a double(14,4),
2005-09-23 10:15:11 +02:00
b varchar(10),
index (a,b)
) engine=merge union=(t2,t3);
create table t2 (
2005-09-27 15:26:16 +02:00
a double(14,4),
2005-09-23 10:15:11 +02:00
b varchar(10),
index (a,b)
) engine=myisam;
create table t3 (
2005-09-27 15:26:16 +02:00
a double(14,4),
2005-09-23 10:15:11 +02:00
b varchar(10),
index (a,b)
) engine=myisam;
insert into t2 values ( null, '');
2005-09-29 13:37:30 +02:00
insert into t2 values ( 9999999999.999, '');
2005-09-23 10:15:11 +02:00
insert into t3 select * from t2;
select min(a), max(a) from t1;
min(a) max(a)
2005-09-29 13:37:30 +02:00
9999999999.9990 9999999999.9990
2005-09-23 10:15:11 +02:00
flush tables;
select min(a), max(a) from t1;
min(a) max(a)
2005-09-29 13:37:30 +02:00
9999999999.9990 9999999999.9990
2005-09-23 10:15:11 +02:00
drop table t1, t2, t3;
2004-11-29 04:51:30 +01:00
create table t1 (a int,b int,c int, index (a,b,c));
create table t2 (a int,b int,c int, index (a,b,c));
create table t3 (a int,b int,c int, index (a,b,c))
engine=merge union=(t1 ,t2);
insert into t1 (a,b,c) values (1,1,0),(1,2,0);
insert into t2 (a,b,c) values (1,1,1),(1,2,1);
explain select a,b,c from t3 force index (a) where a=1 order by a,b,c;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE t3 ref a a 5 const 2 Using where; Using index
select a,b,c from t3 force index (a) where a=1 order by a,b,c;
a b c
1 1 0
1 1 1
1 2 0
1 2 1
explain select a,b,c from t3 force index (a) where a=1 order by a desc, b desc, c desc;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE t3 ref a a 5 const 2 Using where; Using index
select a,b,c from t3 force index (a) where a=1 order by a desc, b desc, c desc;
a b c
1 2 1
1 2 0
1 1 1
1 1 0
2004-12-27 13:05:35 +01:00
show index from t3;
2010-02-20 11:07:32 +01:00
Table Non_unique Key_name Seq_in_index Column_name Collation Cardinality Sub_part Packed Null Index_type Comment Index_comment
t3 1 a 1 a A NULL NULL NULL YES BTREE
t3 1 a 2 b A NULL NULL NULL YES BTREE
t3 1 a 3 c A NULL NULL NULL YES BTREE
2004-11-29 04:51:30 +01:00
drop table t1, t2, t3;
2005-05-18 19:40:39 +02:00
CREATE TABLE t1 ( a INT AUTO_INCREMENT PRIMARY KEY, b VARCHAR(10), UNIQUE (b) )
ENGINE=MyISAM;
CREATE TABLE t2 ( a INT AUTO_INCREMENT, b VARCHAR(10), INDEX (a), INDEX (b) )
ENGINE=MERGE UNION (t1) INSERT_METHOD=FIRST;
INSERT INTO t2 (b) VALUES (1) ON DUPLICATE KEY UPDATE b=2;
INSERT INTO t2 (b) VALUES (1) ON DUPLICATE KEY UPDATE b=3;
SELECT b FROM t2;
b
3
DROP TABLE t1, t2;
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Extended the unique table check by a check of lock data.
Merge sub-tables cannot be detected by doing name checks only.
mysql-test/r/create.result:
BUG#5390 - problems with merge tables
Removed a duplicate test.
mysql-test/r/merge.result:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT
Added test results.
mysql-test/t/create.test:
BUG#5390 - problems with merge tables
Removed a duplicate test.
mysql-test/t/merge.test:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT
Added tests.
sql/lock.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added a new function to find a duplicate lock in a list of tables.
sql/mysql_priv.h:
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
Added a declaration for the new function.
sql/sql_base.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
Added a call to the new mysql_lock_have_duplicate(),
which needs the thread handle, to unique_table().
sql/sql_delete.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
sql/sql_insert.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
sql/sql_load.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
sql/sql_parse.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
sql/sql_update.cc:
BUG#5390 - problems with merge tables
Problem #1: INSERT...SELECT, Version for 5.1.
Added the thread handle to unique_table().
Replaced a call to find_table_in_local_list() by
the newly extended unique_table().
2005-12-22 13:48:00 +01:00
create table t1(a int);
create table t2(a int);
insert into t1 values (1);
insert into t2 values (2);
create table t3 (a int) engine=merge union=(t1, t2) insert_method=first;
select * from t3;
a
1
2
insert t2 select * from t2;
select * from t2;
a
2
2
insert t3 select * from t1;
select * from t3;
a
1
1
2
2
insert t1 select * from t3;
select * from t1;
a
1
1
1
1
2
2
select * from t2;
a
2
2
select * from t3;
a
1
1
1
1
2
2
2
2
check table t1, t2;
Table Op Msg_type Msg_text
test.t1 check status OK
test.t2 check status OK
drop table t1, t2, t3;
2006-09-28 19:10:06 +02:00
CREATE TABLE t1(a INT);
INSERT INTO t1 VALUES(2),(1);
CREATE TABLE t2(a INT, KEY(a)) ENGINE=MERGE UNION=(t1);
SELECT * FROM t2 WHERE a=2;
2007-01-31 13:15:20 +01:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2006-09-28 19:10:06 +02:00
DROP TABLE t1, t2;
2006-09-27 15:33:31 +02:00
CREATE TABLE t1(a INT) ENGINE=MEMORY;
CREATE TABLE t2(a INT) ENGINE=MERGE UNION=(t1);
SELECT * FROM t2;
2006-10-06 07:54:47 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2006-09-27 15:33:31 +02:00
DROP TABLE t1, t2;
CREATE TABLE t2(a INT) ENGINE=MERGE UNION=(t3);
SELECT * FROM t2;
2008-10-09 10:55:16 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2006-09-27 15:33:31 +02:00
DROP TABLE t2;
2007-01-31 13:15:20 +01:00
CREATE TABLE t1(a INT, b TEXT);
CREATE TABLE tm1(a TEXT, b INT) ENGINE=MERGE UNION=(t1);
SELECT * FROM tm1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, tm1;
CREATE TABLE t1(a SMALLINT, b SMALLINT);
CREATE TABLE tm1(a INT) ENGINE=MERGE UNION=(t1);
SELECT * FROM tm1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, tm1;
CREATE TABLE t1(a SMALLINT, b SMALLINT, KEY(a, b));
CREATE TABLE tm1(a SMALLINT, b SMALLINT, KEY(a)) ENGINE=MERGE UNION=(t1);
SELECT * FROM tm1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, tm1;
CREATE TABLE t1(a SMALLINT, b SMALLINT, KEY(b));
CREATE TABLE tm1(a SMALLINT, b SMALLINT, KEY(a)) ENGINE=MERGE UNION=(t1);
SELECT * FROM tm1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, tm1;
BUG#26881 - Large MERGE tables report incorrect specification when no
differences in tables
Certain merge tables were wrongly reported as having incorrect definition:
- Some fields that are 1 byte long (e.g. TINYINT, CHAR(1)), might
be internally casted (in certain cases) to a different type on a
storage engine layer. (affects 4.1 and up)
- If tables in a merge (and a MERGE table itself) had short VARCHAR column (less
than 4 bytes) and at least one (but not all) tables were ALTER'ed (even to an
identical table: ALTER TABLE xxx ENGINE=yyy), table definitions went ouf of
sync. (affects 4.1 only)
This is fixed by relaxing a check for underlying conformance and setting
field type to FIELD_TYPE_STRING in case varchar is shorter than 4
when a table is created.
myisam/mi_create.c:
Added a comment.
mysql-test/r/merge.result:
A test case for bug#26881.
mysql-test/t/merge.test:
A test case for bug#26881.
sql/ha_myisam.cc:
Relaxed some checks performed by check_definition():
As comparing of fulltext keys (and key segments) is not yet implemented,
only return an error in case one of keys is fulltext and other is not.
Otherwise, if both keys are fulltext, accept them as is.
As comparing of spatial keys (and key segments) is not yet implemented,
only return an error in case one of keys is spatial and other is not.
Otherwise, if both keys are spatial, accept them as is.
A workaround to handle situation when field is casted from FIELD_SKIP_ZERO
to FIELD_NORMAL. This could happen only in case field length is 1 and row
format is fixed.
sql/sql_parse.cc:
When a table that has varchar field shorter than 4 is created, field type is
set to FIELD_TYPE_VAR_STRING. Later, when a table is modified using alter
table, field type is changed to FIELD_TYPE_STRING (see Field_string::type).
That means HA_OPTION_PACK_RECORD flag might be lost and thus null_bit might
be shifted by alter table, in other words alter table doesn't create 100%
equal table definition.
This is usually not a problem, since when a table is created/altered,
definition on a storage engine layer is based on one that is passed from
sql layer. But it is a problem for merge engine - null_bit is shifted when
a table (merge or underlying) is altered.
Set field type to FIELD_TYPE_STRING in case FIELD_TYPE_VAR_STRING is shorter
than 4 when a table is created as it is done in Field::type.
2007-03-13 15:02:06 +01:00
CREATE TABLE t1(c1 VARCHAR(1));
CREATE TABLE m1 LIKE t1;
ALTER TABLE m1 ENGINE=MERGE UNION=(t1);
SELECT * FROM m1;
c1
DROP TABLE t1, m1;
CREATE TABLE t1(c1 VARCHAR(4), c2 TINYINT, c3 TINYINT, c4 TINYINT,
c5 TINYINT, c6 TINYINT, c7 TINYINT, c8 TINYINT, c9 TINYINT);
CREATE TABLE m1 LIKE t1;
ALTER TABLE m1 ENGINE=MERGE UNION=(t1);
SELECT * FROM m1;
c1 c2 c3 c4 c5 c6 c7 c8 c9
DROP TABLE t1, m1;
2007-04-10 22:40:35 +02:00
CREATE TABLE t1 (a VARCHAR(255) CHARACTER SET latin1 COLLATE latin1_german2_ci,
b INT, INDEX(a,b));
CREATE TABLE t2 LIKE t1;
CREATE TABLE t3 LIKE t1;
ALTER TABLE t3 ENGINE=MERGE UNION=(t1,t2);
INSERT INTO t1 VALUES ('ss',1);
INSERT INTO t2 VALUES ('ss',2),(0xDF,2);
SELECT COUNT(*) FROM t3 WHERE a=0xDF AND b=2;
COUNT(*)
2
DROP TABLE t1,t2,t3;
2006-05-31 02:10:53 +02:00
create table t1 (b bit(1));
create table t2 (b bit(1));
create table tm (b bit(1)) engine = merge union = (t1,t2);
select * from tm;
b
drop table tm, t1, t2;
2006-07-12 02:28:50 +02:00
create table t1 (a int) insert_method = last engine = merge;
insert into t1 values (1);
ERROR HY000: Table 't1' is read only
create table t2 (a int) engine = myisam;
alter table t1 union (t2);
insert into t1 values (1);
alter table t1 insert_method = no;
insert into t1 values (1);
ERROR HY000: Table 't1' is read only
drop table t2;
drop table t1;
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
CREATE TABLE tm1(a INT) ENGINE=MERGE UNION=(t1, t2);
SELECT * FROM tm1;
2008-10-09 10:55:16 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
CHECK TABLE tm1;
Table Op Msg_type Msg_text
2011-07-22 14:31:10 +02:00
test.tm1 check Error Table 'test.t1' is differently defined or of non-MyISAM type or doesn't exist
test.tm1 check Error Table 'test.t2' is differently defined or of non-MyISAM type or doesn't exist
2008-10-09 10:55:16 +02:00
test.tm1 check Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
test.tm1 check error Corrupt
CREATE TABLE t1(a INT);
SELECT * FROM tm1;
2008-10-09 10:55:16 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
CHECK TABLE tm1;
Table Op Msg_type Msg_text
2011-07-22 14:31:10 +02:00
test.tm1 check Error Table 'test.t2' is differently defined or of non-MyISAM type or doesn't exist
2008-10-09 10:55:16 +02:00
test.tm1 check Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
test.tm1 check error Corrupt
CREATE TABLE t2(a BLOB);
SELECT * FROM tm1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
CHECK TABLE tm1;
Table Op Msg_type Msg_text
2011-07-22 14:31:10 +02:00
test.tm1 check Error Table 'test.t2' is differently defined or of non-MyISAM type or doesn't exist
BUG#26976 - Missing table in merge not noted in related error msg +
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
myisammrg/myrg_def.h:
Added myrg_print_wrong_table declaration.
myisammrg/myrg_open.c:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
mysql-test/r/backup.result:
Updated test result.
mysql-test/r/key_cache.result:
Updated test result - removed duplicate error.
mysql-test/r/lock.result:
Updated test result - added summary row.
mysql-test/r/merge.result:
A test case for BUG#26976.
mysql-test/r/preload.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/ps.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/repair.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/rpl_failed_optimize.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/sp.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/r/view.result:
Updated test result - removed duplicate error, added summary row.
mysql-test/t/merge.test:
A test case for BUG#26976.
sql/ha_myisam.cc:
Do not report same error twice.
sql/ha_myisammrg.cc:
If HA_OPEN_FOR_REPAIR is passed to merge engine open function,
report names of tables that we fail to open.
Added dummy ha_myisammrg::check to not confuse users with
"not implemented" error in case all underlying tables are
fine.
sql/ha_myisammrg.h:
Added ha_myisammrg::check declaration.
sql/share/errmsg.txt:
Added ER_ADMIN_WRONG_MRG_TABLE errno. It is used instead
of ER_WRONG_MRG_TABLE in case HA_OPEN_FOR_REPAIR is
passed to merge engine handler open function.
sql/sql_error.cc:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_error.h:
warning_level_* are now public. It is required by mysql_admin_table to
report message level.
sql/sql_table.cc:
Reorder mysql_admin_table arguments to meet it's definition.
Report errors that are pending in thd->warn_list as results of
admin function.
2007-06-06 01:42:41 +02:00
test.tm1 check Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.tm1 check error Corrupt
ALTER TABLE t2 MODIFY a INT;
SELECT * FROM tm1;
a
CHECK TABLE tm1;
Table Op Msg_type Msg_text
test.tm1 check status OK
DROP TABLE tm1, t1, t2;
2007-10-09 13:10:51 +02:00
CREATE TABLE t1(c1 INT);
CREATE TABLE t2 (c1 INT) ENGINE=MERGE UNION=(t1) INSERT_METHOD=FIRST;
CREATE TABLE IF NOT EXISTS t1 SELECT * FROM t2;
2010-08-18 11:35:41 +02:00
Warnings:
Note 1050 Table 't1' already exists
2007-10-09 13:10:51 +02:00
DROP TABLE t1, t2;
2007-11-26 16:58:54 +01:00
CREATE TABLE t1 (id INT NOT NULL, ref INT NOT NULL, INDEX (id)) ENGINE=MyISAM;
CREATE TABLE t2 LIKE t1;
INSERT INTO t2 (id, ref) VALUES (1,3), (2,1), (3,2), (4,5), (4,4);
INSERT INTO t1 SELECT * FROM t2;
INSERT INTO t1 SELECT * FROM t2;
CREATE TABLE t3 (id INT NOT NULL, ref INT NOT NULL, INDEX (id)) ENGINE=MERGE
UNION(t1);
SELECT * FROM t3 AS a INNER JOIN t3 AS b USING (id) WHERE a.ref < b.ref;
id ref ref
4 4 5
4 4 5
4 4 5
4 4 5
SELECT * FROM t3;
id ref
1 3
2 1
3 2
4 5
4 4
1 3
2 1
3 2
4 5
4 4
DELETE FROM a USING t3 AS a INNER JOIN t3 AS b USING (id) WHERE a.ref < b.ref;
SELECT * FROM t3;
id ref
1 3
2 1
3 2
4 5
1 3
2 1
3 2
4 5
DROP TABLE t1, t2, t3;
2008-03-14 16:38:22 +01:00
CREATE TABLE t1(a INT);
CREATE TABLE m1(a INT) ENGINE=MERGE;
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
2008-03-14 21:37:19 +01:00
`a` int(11) DEFAULT NULL
2008-03-14 16:38:22 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1
DROP TABLE m1;
CREATE TABLE m1(a INT) ENGINE=MERGE UNION=();
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
2008-03-14 21:37:19 +01:00
`a` int(11) DEFAULT NULL
2008-03-14 16:38:22 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1
ALTER TABLE m1 UNION=(t1);
ALTER TABLE m1 UNION=();
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
2008-03-14 21:37:19 +01:00
`a` int(11) DEFAULT NULL
2008-03-14 16:38:22 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1
DROP TABLE t1, m1;
2009-02-04 12:46:23 +01:00
CREATE TABLE t1(a INT);
CREATE TABLE t2(a VARCHAR(10));
CREATE TABLE m1(a INT) ENGINE=MERGE UNION=(t1, t2);
CREATE TABLE m2(a INT) ENGINE=MERGE UNION=(t1);
SELECT * FROM t1;
a
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
SELECT * FROM m2;
a
DROP TABLE t1, t2, m1, m2;
2006-07-12 02:28:50 +02:00
End of 5.0 tests
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
create table t1 (c1 int, index(c1));
create table t2 (c1 int, index(c1)) engine=merge union=(t1);
insert into t1 values (1);
flush tables;
select * from t2;
c1
1
flush tables;
truncate table t1;
insert into t1 values (1);
flush tables;
select * from t2;
c1
1
truncate table t1;
insert into t1 values (1);
drop table t1,t2;
#
# Extra tests for TRUNCATE.
#
# Truncate MERGE table.
CREATE TABLE t1 (c1 INT, INDEX(c1));
CREATE TABLE t2 (c1 INT, INDEX(c1));
CREATE TABLE t3 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1,t2);
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
SELECT * FROM t3;
c1
1
2
TRUNCATE TABLE t3;
SELECT * FROM t3;
c1
#
# Truncate child table.
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
TRUNCATE TABLE t1;
SELECT * FROM t3;
c1
2
#
# Truncate MERGE table under locked tables.
LOCK TABLE t1 WRITE, t2 WRITE, t3 WRITE;
INSERT INTO t1 VALUES (1);
TRUNCATE TABLE t3;
SELECT * FROM t3;
c1
2009-12-11 13:24:23 +01:00
UNLOCK TABLES;
SELECT * FROM t1;
c1
SELECT * FROM t2;
c1
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Truncate child table under locked tables.
2009-12-11 13:24:23 +01:00
LOCK TABLE t1 WRITE, t2 WRITE, t3 WRITE;
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
TRUNCATE TABLE t1;
SELECT * FROM t3;
c1
2
UNLOCK TABLES;
DROP TABLE t1, t2, t3;
#
# Truncate temporary MERGE table.
CREATE TEMPORARY TABLE t1 (c1 INT, INDEX(c1));
CREATE TEMPORARY TABLE t2 (c1 INT, INDEX(c1));
CREATE TEMPORARY TABLE t3 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1,t2);
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
SELECT * FROM t3;
c1
1
2
TRUNCATE TABLE t3;
SELECT * FROM t3;
c1
#
# Truncate temporary child table.
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
TRUNCATE TABLE t1;
SELECT * FROM t3;
c1
2
#
# Truncate temporary MERGE table under locked tables.
INSERT INTO t1 VALUES (1);
CREATE TABLE t4 (c1 INT, INDEX(c1));
LOCK TABLE t4 WRITE;
TRUNCATE TABLE t3;
SELECT * FROM t3;
c1
2009-12-11 13:24:23 +01:00
SELECT * FROM t1;
c1
SELECT * FROM t2;
c1
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Truncate temporary child table under locked tables.
2009-12-11 13:24:23 +01:00
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
TRUNCATE TABLE t1;
SELECT * FROM t3;
c1
2009-12-11 13:24:23 +01:00
2
SELECT * FROM t1;
c1
SELECT * FROM t2;
c1
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
2
UNLOCK TABLES;
DROP TABLE t1, t2, t3, t4;
CREATE TABLE t1 (c1 INT) ENGINE= MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE= MRG_MYISAM UNION= (t1) INSERT_METHOD= LAST;
REPAIR TABLE t1;
INSERT INTO t2 VALUES (1);
Table Op Msg_type Msg_text
test.t1 repair status OK
DROP TABLE t1, t2;
CREATE TABLE t1 (c1 INT) ENGINE= MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE= MRG_MYISAM UNION= (t1) INSERT_METHOD= LAST;
LOCK TABLE t1 WRITE;
INSERT INTO t2 VALUES (1);
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair status OK
UNLOCK TABLES;
DROP TABLE t1, t2;
CREATE TABLE t1 (c1 INT) ENGINE= MyISAM;
LOCK TABLE t1 WRITE;
INSERT INTO t1 VALUES (1);
FLUSH TABLES;
FLUSH TABLES;
SELECT * FROM t1;
c1
UNLOCK TABLES;
DROP TABLE t1;
#
# Extra tests for Bug#26379 - Combination of FLUSH TABLE and
# REPAIR TABLE corrupts a MERGE table
#
# CREATE ... SELECT is disabled for MERGE tables.
#
CREATE TABLE t1(c1 INT);
INSERT INTO t1 VALUES (1);
CREATE TABLE t2 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
CREATE TABLE t3 ENGINE=MRG_MYISAM INSERT_METHOD=LAST SELECT * FROM t2;
ERROR HY000: Table 't3' is read only
SHOW CREATE TABLE t3;
ERROR 42S02: Table 'test.t3' doesn't exist
CREATE TABLE t3 ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST
SELECT * FROM t2;
ERROR HY000: 'test.t3' is not BASE TABLE
SHOW CREATE TABLE t3;
ERROR 42S02: Table 'test.t3' doesn't exist
DROP TABLE t1, t2;
#
2009-12-10 11:53:20 +01:00
# Bug#37371 "CREATE TABLE LIKE merge loses UNION parameter"
# Demonstrate that this is no longer the case.
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# 1. Create like.
CREATE TABLE t1 (c1 INT);
CREATE TABLE t2 (c1 INT);
CREATE TABLE t3 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
INSERT INTO t3 VALUES (3);
CREATE TABLE t4 LIKE t3;
SHOW CREATE TABLE t4;
Table Create Table
t4 CREATE TABLE `t4` (
`c1` int(11) DEFAULT NULL
2009-12-10 11:53:20 +01:00
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
INSERT INTO t4 VALUES (4);
DROP TABLE t4;
#
# 1. Create like with locked tables.
LOCK TABLES t3 WRITE, t2 WRITE, t1 WRITE;
CREATE TABLE t4 LIKE t3;
2009-12-10 11:53:20 +01:00
ERROR HY000: Table 't4' was not locked with LOCK TABLES
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SHOW CREATE TABLE t4;
ERROR HY000: Table 't4' was not locked with LOCK TABLES
INSERT INTO t4 VALUES (4);
ERROR HY000: Table 't4' was not locked with LOCK TABLES
2010-06-25 17:32:59 +02:00
# Temporary tables can be created in spite of LOCK TABLES.
# If the temporary MERGE table uses the locked children only,
# it can even be used.
2009-12-10 11:53:20 +01:00
CREATE TEMPORARY TABLE t4 LIKE t3;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SHOW CREATE TABLE t4;
2010-07-02 18:07:57 +02:00
Table Create Table
t4 CREATE TEMPORARY TABLE `t4` (
`c1` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
INSERT INTO t4 VALUES (4);
2009-12-10 11:53:20 +01:00
UNLOCK TABLES;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
INSERT INTO t4 VALUES (4);
DROP TABLE t4;
#
# Rename child.
#
# 1. Normal rename of non-MERGE table.
CREATE TABLE t4 (c1 INT);
INSERT INTO t4 VALUES (4);
SELECT * FROM t4 ORDER BY c1;
c1
4
RENAME TABLE t4 TO t5;
SELECT * FROM t5 ORDER BY c1;
c1
4
RENAME TABLE t5 TO t4;
SELECT * FROM t4 ORDER BY c1;
c1
4
DROP TABLE t4;
#
# 2. Normal rename.
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t2 TO t5;
SELECT * FROM t3 ORDER BY c1;
2008-10-09 10:55:16 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t5 TO t2;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# 3. Normal rename with locked tables.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t2 TO t5;
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t5 TO t2;
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
UNLOCK TABLES;
#
# 4. Alter table rename.
ALTER TABLE t2 RENAME TO t5;
SELECT * FROM t3 ORDER BY c1;
2008-10-09 10:55:16 +02:00
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
ALTER TABLE t5 RENAME TO t2;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# 5. Alter table rename with locked tables.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE;
ALTER TABLE t2 RENAME TO t5;
SELECT * FROM t3 ORDER BY c1;
2009-12-03 00:09:22 +01:00
ERROR HY000: Table 't2' was not locked with LOCK TABLES
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
ALTER TABLE t5 RENAME TO t2;
ERROR HY000: Table 't5' was not locked with LOCK TABLES
UNLOCK TABLES;
ALTER TABLE t5 RENAME TO t2;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Rename parent.
#
# 1. Normal rename with locked tables.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t3 TO t5;
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
RENAME TABLE t5 TO t3;
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# 5. Alter table rename with locked tables.
ALTER TABLE t3 RENAME TO t5;
SELECT * FROM t5 ORDER BY c1;
ERROR HY000: Table 't5' was not locked with LOCK TABLES
ALTER TABLE t5 RENAME TO t3;
ERROR HY000: Table 't5' was not locked with LOCK TABLES
UNLOCK TABLES;
ALTER TABLE t5 RENAME TO t3;
SELECT * FROM t3 ORDER BY c1;
c1
1
2
3
2009-12-10 11:53:20 +01:00
4
2010-07-02 18:07:57 +02:00
4
4
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
DROP TABLE t1, t2, t3;
#
# Drop locked tables.
#
# 1. Drop parent.
CREATE TABLE t1 (c1 INT, INDEX(c1));
CREATE TABLE t2 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
LOCK TABLES t1 WRITE, t2 WRITE;
INSERT INTO t1 VALUES (1);
DROP TABLE t2;
SELECT * FROM t2;
ERROR HY000: Table 't2' was not locked with LOCK TABLES
SELECT * FROM t1;
c1
1
UNLOCK TABLES;
# 2. Drop child.
CREATE TABLE t2 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
LOCK TABLES t1 WRITE, t2 WRITE;
INSERT INTO t1 VALUES (1);
DROP TABLE t1;
SELECT * FROM t2;
2009-12-03 00:09:22 +01:00
ERROR HY000: Table 't1' was not locked with LOCK TABLES
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SELECT * FROM t1;
2009-12-03 00:09:22 +01:00
ERROR HY000: Table 't1' was not locked with LOCK TABLES
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
UNLOCK TABLES;
DROP TABLE t2;
#
# ALTER TABLE. Change child list.
#
CREATE TABLE t1 (c1 INT, INDEX(c1));
CREATE TABLE t2 (c1 INT, INDEX(c1));
CREATE TABLE t3 (c1 INT, INDEX(c1));
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
INSERT INTO t3 VALUES (3);
CREATE TABLE t4 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t3,t2)
INSERT_METHOD=LAST;
# Shrink child list.
ALTER TABLE t4 UNION=(t3);
SHOW CREATE TABLE t4;
Table Create Table
t4 CREATE TABLE `t4` (
`c1` int(11) DEFAULT NULL,
KEY `c1` (`c1`)
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t3`)
SELECT * FROM t4 ORDER BY c1;
c1
3
# Extend child list.
ALTER TABLE t4 UNION=(t3,t2);
SHOW CREATE TABLE t4;
Table Create Table
t4 CREATE TABLE `t4` (
`c1` int(11) DEFAULT NULL,
KEY `c1` (`c1`)
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t3`,`t2`)
SELECT * FROM t4 ORDER BY c1;
c1
2
3
#
# ALTER TABLE under LOCK TABLES. Change child list.
#
LOCK TABLES t4 WRITE, t3 WRITE, t2 WRITE;
# Shrink child list.
ALTER TABLE t4 UNION=(t3);
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
# Extend child list within locked tables.
ALTER TABLE t4 UNION=(t3,t2);
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
# Extend child list beyond locked tables.
ALTER TABLE t4 UNION=(t3,t2,t1);
ERROR HY000: Can't execute the given command because you have active locked tables or an active transaction
SHOW CREATE TABLE t4;
Table Create Table
t4 CREATE TABLE `t4` (
`c1` int(11) DEFAULT NULL,
KEY `c1` (`c1`)
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t3`,`t2`)
SELECT * FROM t4 ORDER BY c1;
c1
2
3
UNLOCK TABLES;
DROP TABLE t4;
#
# ALTER TABLE under LOCK TABLES. Grave change, table re-creation.
#
CREATE TABLE t4 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1,t2,t3)
INSERT_METHOD=LAST;
# Lock parent first and then children.
LOCK TABLES t4 WRITE, t3 WRITE, t2 WRITE, t1 WRITE;
ALTER TABLE t4 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock children first and then parent.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE, t4 WRITE;
ALTER TABLE t4 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock parent between children.
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
ALTER TABLE t4 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 DROP INDEX c1, ADD UNIQUE INDEX (c1);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
DROP TABLE t1, t2, t3, t4;
#
# ALTER TABLE under LOCK TABLES. Simple change, no re-creation.
#
CREATE TABLE t1 (c1 INT);
CREATE TABLE t2 (c1 INT);
CREATE TABLE t3 (c1 INT);
CREATE TABLE t4 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2,t3)
INSERT_METHOD=LAST;
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
INSERT INTO t3 VALUES (3);
# Lock parent first and then children.
LOCK TABLES t4 WRITE, t3 WRITE, t2 WRITE, t1 WRITE;
ALTER TABLE t4 ALTER COLUMN c1 SET DEFAULT 44;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 ALTER COLUMN c1 SET DEFAULT 22;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock children first and then parent.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE, t4 WRITE;
ALTER TABLE t4 ALTER COLUMN c1 SET DEFAULT 44;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 ALTER COLUMN c1 SET DEFAULT 22;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock parent between children.
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
ALTER TABLE t4 ALTER COLUMN c1 SET DEFAULT 44;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
ALTER TABLE t2 ALTER COLUMN c1 SET DEFAULT 22;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
#
# FLUSH TABLE under LOCK TABLES.
#
# Lock parent first and then children.
LOCK TABLES t4 WRITE, t3 WRITE, t2 WRITE, t1 WRITE;
FLUSH TABLE t4;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLE t2;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLES;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock children first and then parent.
LOCK TABLES t1 WRITE, t2 WRITE, t3 WRITE, t4 WRITE;
FLUSH TABLE t4;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLE t2;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLES;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
# Lock parent between children.
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
FLUSH TABLE t4;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLE t2;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
FLUSH TABLES;
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
UNLOCK TABLES;
#
# Triggers
#
# Trigger on parent
DELETE FROM t4 WHERE c1 = 4;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t4_ai1 AFTER INSERT ON t4 FOR EACH ROW SET @a=1;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SET @a=0;
INSERT INTO t4 VALUES (4);
SELECT @a;
@a
1
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t4_ai1;
CHECK TABLE t3;
Table Op Msg_type Msg_text
test.t3 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
# Trigger on parent under LOCK TABLES
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t4_ai2 AFTER INSERT ON t4 FOR EACH ROW SET @a=1;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SET @a=0;
INSERT INTO t4 VALUES (4);
SELECT @a;
@a
1
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
4
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t4_ai2;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
UNLOCK TABLES;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CHECK TABLE t3;
Table Op Msg_type Msg_text
test.t3 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Trigger on child
DELETE FROM t4 WHERE c1 = 4;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t3_ai3 AFTER INSERT ON t3 FOR EACH ROW SET @a=1;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SET @a=0;
INSERT INTO t4 VALUES (4);
SELECT @a;
@a
0
INSERT INTO t3 VALUES (33);
SELECT @a;
@a
1
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
33
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t3_ai3;
CHECK TABLE t3;
Table Op Msg_type Msg_text
test.t3 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
# Trigger on child under LOCK TABLES
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t3_ai4 AFTER INSERT ON t3 FOR EACH ROW SET @a=1;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
SET @a=0;
INSERT INTO t4 VALUES (4);
SELECT @a;
@a
0
INSERT INTO t3 VALUES (33);
SELECT @a;
@a
1
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
4
33
33
DELETE FROM t4 WHERE c1 = 33;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t3_ai4;
2010-06-25 17:32:59 +02:00
UNLOCK TABLES;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CHECK TABLE t3;
Table Op Msg_type Msg_text
test.t3 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Trigger with table use on child
DELETE FROM t4 WHERE c1 = 4;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t3_ai5 AFTER INSERT ON t3 FOR EACH ROW INSERT INTO t2 VALUES(22);
SELECT COUNT(*) FROM t2;
COUNT(*)
1
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
INSERT INTO t4 VALUES (4);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
INSERT INTO t3 VALUES (33);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
22
33
DELETE FROM t4 WHERE c1 = 22;
DELETE FROM t4 WHERE c1 = 33;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t3_ai5;
UNLOCK TABLES;
CHECK TABLE t2,t3;
Table Op Msg_type Msg_text
test.t2 check status OK
test.t3 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
# Trigger with table use on child under LOCK TABLES
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
CREATE TRIGGER t3_ai6 AFTER INSERT ON t3 FOR EACH ROW INSERT INTO t2 VALUES(22);
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
INSERT INTO t4 VALUES (4);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
4
INSERT INTO t3 VALUES (33);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
4
22
33
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
DROP TRIGGER t3_ai6;
UNLOCK TABLES;
check table t2,t3,t4;
Table Op Msg_type Msg_text
test.t2 check status OK
test.t3 check status OK
test.t4 check status OK
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
DELETE FROM t4 WHERE c1 = 22;
DELETE FROM t4 WHERE c1 = 33;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
# Trigger with table use on child under different LOCK TABLES
DELETE FROM t4 WHERE c1 = 4;
LOCK TABLES t4 WRITE,t3 WRITE, t2 WRITE, t1 WRITE;
CREATE TRIGGER t3_ai7 AFTER INSERT ON t3 FOR EACH ROW INSERT INTO t2 VALUES(22);
INSERT INTO t4 VALUES (4);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
INSERT INTO t3 VALUES (33);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
22
33
DROP TRIGGER t3_ai7;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
UNLOCK TABLES;
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
check table t2,t3,t4;
Table Op Msg_type Msg_text
test.t2 check status OK
test.t3 check status OK
test.t4 check status OK
DELETE FROM t4 WHERE c1 = 22;
DELETE FROM t4 WHERE c1 = 33;
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
#
# Repair
#
REPAIR TABLE t4;
Table Op Msg_type Msg_text
test.t4 repair note The storage engine for the table doesn't support repair
REPAIR TABLE t2;
Table Op Msg_type Msg_text
test.t2 repair status OK
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
REPAIR TABLE t4;
Table Op Msg_type Msg_text
test.t4 repair note The storage engine for the table doesn't support repair
REPAIR TABLE t2;
Table Op Msg_type Msg_text
test.t2 repair status OK
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
UNLOCK TABLES;
#
# Optimize
#
OPTIMIZE TABLE t4;
Table Op Msg_type Msg_text
test.t4 optimize note The storage engine for the table doesn't support optimize
OPTIMIZE TABLE t2;
Table Op Msg_type Msg_text
test.t2 optimize status OK
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
OPTIMIZE TABLE t4;
Table Op Msg_type Msg_text
test.t4 optimize note The storage engine for the table doesn't support optimize
OPTIMIZE TABLE t2;
Table Op Msg_type Msg_text
test.t2 optimize status Table is already up to date
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
UNLOCK TABLES;
#
# Checksum
#
CHECKSUM TABLE t4;
Table Checksum
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
test.t4 149057747
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
CHECKSUM TABLE t2;
Table Checksum
test.t2 3700403066
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
CHECKSUM TABLE t4;
Table Checksum
Fix for: LP #634955: Assert in _ma_update_at_original_place()
Added locking of lock mutex when updating status in external_unlock() for Aria and MyISAM tables.
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards. (Allows one to run --debug with mysql-test-run scripts that uses @debug)
Fixed several core dump, deadlock and crashed table bugs in handling of LOCK TABLE with MERGE tables:
- Added priority of locks to avoid crashes with MERGE tables.
- Added thr_lock_merge() to allow one to merge two results of thr_lock().
Fixed 'not found row' bug in REPLACE with Aria tables.
Mark MyISAM tables that are part of MERGE with HA_OPEN_MERGE_TABLE and set the locks to have priority THR_LOCK_MERGE_PRIV.
- By sorting MERGE tables last in thr_multi_unlock() it's safer to release and relock them many times (can happen when TRIGGERS are created)
Avoid printing (null) in debug file (to easier find out wrong NULL pointer usage with %s).
client/mysql.cc:
Fixed that 'source' command doesn't cause mysql command line tool to exit on error.
client/mysqltest.cc:
Don't send NULL to fn_format(). (Can cause crash on Solaris when using --debug)
dbug/dbug.c:
DEBUG_EXECUTE() and DEBUG_EVALUATE_IF() should not execute things based on wildcards.
include/my_base.h:
Added flag to signal if one opens a MERGE table.
Added extra() command to signal that one is not part of a MERGE table anymore.
include/thr_lock.h:
Added priority for locks (needed to fix bug in thr_lock when using MERGE tables)
Added option to thr_unlock() if get_status() should be called.
Added prototype for thr_merge_locks().
mysql-test/mysql-test-run.pl:
Ignore crashed table warnings for tables named 'crashed'.
mysql-test/r/merge.result:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/r/merge_debug.result:
Test of error handling when reopening MERGE tables.
mysql-test/r/udf_query_cache.result:
Added missing flush status
mysql-test/suite/parts/r/partition_repair_myisam.result:
Update results
mysql-test/t/merge.test:
Renamed triggers to make debugging easier.
Added some CHECK TABLES to catch errors earlier.
Additional tests.
mysql-test/t/merge_debug.test:
Test of error handling when reopening MERGE tables.
mysql-test/t/udf_query_cache.test:
Added missing flush status
mysys/my_getopt.c:
Removed not used variable
mysys/my_symlink2.c:
Changed (null) to (NULL) to make it easier to find NULL arguments to DBUG_PRINT() functions.
(On linux, NULL to sprintf is printed 'null')
mysys/thr_lock.c:
Added priority of locks to avoid crashes with MERGE tables.
Added thr_lock_merge() to allow one to merge two results of thr_lock().
- This is needed for MyISAM as all locked table must share the same status. If not, you will not see newly inserted rows in other instances of the table.
If calling thr_unlock() with THR_UNLOCK_UPDATE_STATUS, call update_status() and restore_status() for the locks. This is needed in some rare cases where we call thr_unlock() followed by thr_lock() without calling external_unlock/external_lock in between.
Simplify loop in thr_multi_lock().
Added 'start_trans', which is called at end of thr_multi_lock() when all locks are taken.
- This was needed by Aria to ensure that transaction is started when we got all locks, not at get_status(). Without this, some rows could not be visible when we lock two tables at the same time, causing REPLACE using two tables to fail unexpectedly.
sql/handler.cc:
Add an assert() in handler::print_error() for "impossible errors" (like table is crashed) when --debug-assert-if-crashed-table is used.
sql/lock.cc:
Simplify mysql_lock_tables() code if get_lock_data() returns 0 locks.
Added new parameter to thr_multi_unlock()
In mysql_unlock_read_tables(), call first externa_unlock(), then thr_multi_unlock(); This is same order as we do in mysql_unlock_tables().
Don't abort locks in mysql_lock_abort() for merged tables when a MERGE table is deleted; Would cause a spin lock.
Added call to thr_merge_locks() in mysql_lock_merge() to ensure consistency in thr_locks().
- New locks of same type and table is stored after the old lock to ensure that we get the status from the original lock.
sql/mysql_priv.h:
Added debug_assert_if_crashed_table
sql/mysqld.cc:
Added --debug-assert-if-crashed-table
sql/parse_file.cc:
Don't print '(null)' in DBUG_PRINT of no dir given
sql/set_var.cc:
Increase default size of buffer for @debug variable.
sql/sql_base.cc:
In case of error from reopen_table() in reopen_tables(), call unlock_open_table() and restart loop.
- This fixed bug when we twice deleted same table from open_cache.
Don't take name lock for already name locked table in open_unireg_entry().
- Fixed bug when doing repair in reopen_table().
- In detach_merge_children(), always detach if 'clear_refs' is given. We can't trust parent->children_attached as this function can be called twice, first time with clear_refs set to 0.
sql/sql_class.cc:
Changed printing of (null) to "" in set_thd_proc_info()
sql/sql_parse.cc:
Added DBUG
sql/sql_trigger.cc:
Don't call unlink_open_table() if reopen_table() fails as the table may already be freed.
storage/maria/ma_bitmap.c:
Fixed DBUG_ASSERT() in allocate_tail()
storage/maria/ma_blockrec.c:
Fixed wrong calculation of row length for very small rows in undo_row_update().
- Fixes ASSERT() when doing undo.
storage/maria/ma_blockrec.h:
Added _ma_block_start_trans() and _ma_block_start_trans_no_versioning()
storage/maria/ma_locking.c:
Call _ma_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
storage/maria/ma_open.c:
Changed to use start_trans() instead of get_status() to ensure that we see all rows in all locked tables when we got the locks.
- Fixed 'not found row' bug in REPLACE with Aria tables.
storage/maria/ma_state.c:
Added _ma_update_status_with_lock() and _ma_block_start_trans().
This is to ensure that we see all rows in all locked tables when we got the locks.
storage/maria/ma_state.h:
Added _ma_update_status_with_lock()
storage/maria/ma_write.c:
More DBUG_PRINT
storage/myisam/mi_check.c:
Fixed error message
storage/myisam/mi_extra.c:
Added HA_EXTRA_DETACH_CHILD:
- Detach MyISAM table to not be part of MERGE table (remove flag & lock priority).
storage/myisam/mi_locking.c:
Call mi_update_status_with_lock() when releasing write locks.
- Fixes potential problem with updating status without the proper lock.
Change to use new HA_OPEN_MERGE_TABLE flag to test if MERGE table.
Added mi_fix_status(), called by thr_merge().
storage/myisam/mi_open.c:
Added marker if part of MERGE table.
Call mi_fix_status() in thr_lock() for transactional tables.
storage/myisam/myisamdef.h:
Change my_once_flag to uint, as it stored different values than just 0/1
Added 'open_flag' to store state given to mi_open()
storage/myisammrg/ha_myisammrg.cc:
Add THR_LOCK_MERGE_PRIV to THR_LOCK_DATA to get MERGE locks sorted after other types of locks.
storage/myisammrg/myrg_locking.c:
Remove windows specific code.
storage/myisammrg/myrg_open.c:
Use HA_OPEN_MERGE_TABLE to mi_open().
Set HA_OPEN_MERGE_TABLE for linked MyISAM tables.
storage/xtradb/buf/buf0buf.c:
Fixed compiler warning
storage/xtradb/buf/buf0lru.c:
Initialize variable that could be used not initialized.
2010-11-02 16:22:57 +01:00
test.t4 149057747
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Bug 25038 - Waiting TRUNCATE
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Bug 19627 - temporary merge table locking
Bug 27660 - Falcon: merge table possible
Bug 30273 - merge tables: Can't lock file (errno: 155)
The problems were:
Bug 26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
1. A thread trying to lock a MERGE table performs busy waiting while
REPAIR TABLE or a similar table administration task is ongoing on
one or more of its MyISAM tables.
2. A thread trying to lock a MERGE table performs busy waiting until all
threads that did REPAIR TABLE or similar table administration tasks
on one or more of its MyISAM tables in LOCK TABLES segments do UNLOCK
TABLES. The difference against problem #1 is that the busy waiting
takes place *after* the administration task. It is terminated by
UNLOCK TABLES only.
3. Two FLUSH TABLES within a LOCK TABLES segment can invalidate the
lock. This does *not* require a MERGE table. The first FLUSH TABLES
can be replaced by any statement that requires other threads to
reopen the table. In 5.0 and 5.1 a single FLUSH TABLES can provoke
the problem.
Bug 26867 - LOCK TABLES + REPAIR + merge table result in
memory/cpu hogging
Trying DML on a MERGE table, which has a child locked and
repaired by another thread, made an infinite loop in the server.
Bug 26377 - Deadlock with MERGE and FLUSH TABLE
Locking a MERGE table and its children in parent-child order
and flushing the child deadlocked the server.
Bug 25038 - Waiting TRUNCATE
Truncating a MERGE child, while the MERGE table was in use,
let the truncate fail instead of waiting for the table to
become free.
Bug 25700 - merge base tables get corrupted by
optimize/analyze/repair table
Repairing a child of an open MERGE table corrupted the child.
It was necessary to FLUSH the child first.
Bug 30275 - Merge tables: flush tables or unlock tables
causes server to crash
Flushing and optimizing locked MERGE children crashed the server.
Bug 19627 - temporary merge table locking
Use of a temporary MERGE table with non-temporary children
could corrupt the children.
Temporary tables are never locked. So we do now prohibit
non-temporary chidlren of a temporary MERGE table.
Bug 27660 - Falcon: merge table possible
It was possible to create a MERGE table with non-MyISAM children.
Bug 30273 - merge tables: Can't lock file (errno: 155)
This was a Windows-only bug. Table administration statements
sometimes failed with "Can't lock file (errno: 155)".
These bugs are fixed by a new implementation of MERGE table open.
When opening a MERGE table in open_tables() we do now add the
child tables to the list of tables to be opened by open_tables()
(the "query_list"). The children are not opened in the handler at
this stage.
After opening the parent, open_tables() opens each child from the
now extended query_list. When the last child is opened, we remove
the children from the query_list again and attach the children to
the parent. This behaves similar to the old open. However it does
not open the MyISAM tables directly, but grabs them from the already
open children.
When closing a MERGE table in close_thread_table() we detach the
children only. Closing of the children is done implicitly because
they are in thd->open_tables.
For more detail see the comment at the top of ha_myisammrg.cc.
Changed from open_ltable() to open_and_lock_tables() in all places
that can be relevant for MERGE tables. The latter can handle tables
added to the list on the fly. When open_ltable() was used in a loop
over a list of tables, the list must be temporarily terminated
after every table for open_and_lock_tables().
table_list->required_type is set to FRMTYPE_TABLE to avoid open of
special tables. Handling of derived tables is suppressed.
These details are handled by the new function
open_n_lock_single_table(), which has nearly the same signature as
open_ltable() and can replace it in most cases.
In reopen_tables() some of the tables open by a thread can be
closed and reopened. When a MERGE child is affected, the parent
must be closed and reopened too. Closing of the parent is forced
before the first child is closed. Reopen happens in the order of
thd->open_tables. MERGE parents do not attach their children
automatically at open. This is done after all tables are reopened.
So all children are open when attaching them.
Special lock handling like mysql_lock_abort() or mysql_lock_remove()
needs to be suppressed for MERGE children or forwarded to the parent.
This depends on the situation. In loops over all open tables one
suppresses child lock handling. When a single table is touched,
forwarding is done.
Behavioral changes:
===================
This patch changes the behavior of temporary MERGE tables.
Temporary MERGE must have temporary children.
The old behavior was wrong. A temporary table is not locked. Hence
even non-temporary children were not locked. See
Bug 19627 - temporary merge table locking.
You cannot change the union list of a non-temporary MERGE table
when LOCK TABLES is in effect. The following does *not* work:
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ...;
LOCK TABLES t1 WRITE, t2 WRITE, m1 WRITE;
ALTER TABLE m1 ... UNION=(t1,t2) ...;
However, you can do this with a temporary MERGE table.
You cannot create a MERGE table with CREATE ... SELECT, neither
as a temporary MERGE table, nor as a non-temporary MERGE table.
CREATE TABLE m1 ... ENGINE=MRG_MYISAM ... SELECT ...;
Gives error message: table is not BASE TABLE.
include/my_base.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN.
include/myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added element 'children_attached' to MYRG_INFO.
Added declarations for myrg_parent_open(),
myrg_attach_children() and myrg_detach_children()
for the new MERGE table open approach.
mysql-test/extra/binlog_tests/blackhole.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Preliminarily added new error message with a comment.
mysql-test/r/create.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/r/delayed.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result from here to merge.result.
mysql-test/r/merge.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed/added test result.
mysql-test/r/myisam.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test result for bug 8306 from here to merge.result.
mysql-test/suite/binlog/r/binlog_stm_blackhole.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test result.
mysql-test/t/create.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed error number.
mysql-test/t/delayed.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test from here to merge.test.
mysql-test/t/merge.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed test for new temporary MERGE table behavior.
Exchanged error numbers by symbolic codes.
Added tests. Included are tests for bugs
8306 (moved from myisam.test), 26379, 19627, 25038, 25700, 26377,
26867, 27660, 30275, and 30273.
Fixed changes resulting from disabled CREATE...SELECT.
Integrated tests moved from delayed.test and myisam.test to here.
mysql-test/t/myisam.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Moved test for bug 8306 from here to merge.test.
mysys/thr_lock.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code to let the owner of a high priority lock (TL_WRITE_ONLY)
to bypass its own lock.
sql/ha_ndbcluster_binlog.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/handler.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/mysql_priv.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Removed declaration of check_merge_table_access(). It is now static
in sql_parse.cc.
Added declaration for fix_merge_after_open().
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'.
Added inline functions simple_open_n_lock_tables() and
open_and_lock_tables(), which call open_and_lock_tables_derived()
and add the argument for 'derived'.
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Added 'thd' argument to init_tmp_table_share().
sql/slave.cc:
ug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added comment.
sql/sql_base.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Defined new functions add_merge_table_list(),
attach_merge_children(), detach_merge_children(), and
fix_merge_after_open() for the new MERGE table open approach.
Added calls of the new functions to
close_handle_and_leave_table_as_lock(), close_thread_tables(),
close_thread_table(), unlink_open_table(), reopen_name_locked_table(),
reopen_table(), drop_locked_tables(), close_temporary_table(),
and open_tables() respectively.
Prevented special lock handling of merge children (like
mysql_lock_remove, mysql_lock_merge or mysql_lock_abort)
at many places. Some of these calls are forwarded to the
parent table instead.
Added code to set thd->some_tables_deleted for every thread that has
a table open that we are flushing.
Added code for MERGE tables to unlink_open_table().
Added MERGE children to the list of unusable tables in open_table().
Added MERGE table handling to reopen_table().
Added lock handling and closing of a parent before the children
in close_data_files_and_morph_locks().
Added code for re-attaching children in reopen_tables().
Added MYSQL_LOCK_NOTIFY_IF_NEED_REOPEN to the locking flags and
error reporting after mysql_lock_tables() in reopen_tables().
Added lock handling and closing of a parent before the children
in close_old_data_files().
Added lock handling and detaching in drop_locked_tables().
Added code for removing the children list from the statement list
to prepare for a repetition in open_tables().
Added new function open_n_lock_single_table(), which can be used
as an replacement for open_ltable() in most situations. Internally
it calls simple_open_n_lock_tables() so hat it is appropriate for
MERGE tables.
Disabled use of open_ltable() for MERGE tables.
Removed function simple_open_n_lock_tables(). It is now inline
declared in mysql_priv.h.
Renamed open_and_lock_tables() to open_and_lock_tables_derived()
with additional parameter 'derived'. open_and_lock_tables() is now
inline declared in mysql_priv.h.
Added a check for end-of-list in two loops in lock_tables().
Added 'thd' argument to init_tmp_table_share().
sql/sql_insert.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Changed from open_ltable() to open_n_lock_single_table() in
handle_delayed_insert().
Reestablished LEX settings after lex initialization.
Added 'thd' argument to init_tmp_table_share().
sql/sql_parse.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Made check_merge_table_access() a static function.
Disabled use of CREATE...SELECT for MERGE tables.
sql/sql_partition.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Fixed comment typo.
sql/sql_select.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
sql/sql_table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Optimized use of mysql_ha_flush() in mysql_rm_table_part2().
Disabled the use of MERGE tables with prepare_for_restore() and
prepare_for_repair().
Changed from open_ltable() to open_n_lock_single_table() in
mysql_alter_table() and mysql_checksum_table().
Disabled change of child list under LOCK TABLES.
Initialized table_list->table in mysql_recreate_table().
sql/sql_trigger.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added code for allowing CREATE TRIGGER under LOCK TABLE, to be able
to test it with MERGE tables.
sql/table.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added 'thd' argument to init_tmp_table_share().
Setting table_map_id from query_id in init_tmp_table_share().
Added member function TABLE::is_children_attached().
sql/table.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added access method get_table_def_version() to TABLE_SHARE.
Added elements for MERGE tables to TABLE and TABLE_LIST.
storage/myisam/ha_myisam.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added an unrelated comment to the function comment of table2myisam().
storage/myisam/ha_myisam.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added new member function MI_INFO::file_ptr().
storage/myisammrg/ha_myisammrg.cc:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added callback functions to support parent open and children attach
of MERGE tables.
Changed ha_myisammrg::open() to initialize storage engine structures
and create a list of child tables only. Child tables are not opened.
Added ha_myisammrg::attach_children(), which does now the main part
of MERGE open.
Added ha_myisammrg::detach_children().
Added calls to ::attach_children() and ::detach_children() to
::extra() on HA_EXTRA_ATTACH_CHILDREN and HA_EXTRA_DETACH_CHILDREN
respectively.
Added a check for matching TEMPORARY type for children against
parent.
Added a check for table def version.
Added support for thd->open_options to attach_children().
Changed child path name generation for temporary tables so that
it does nothing special for temporary tables.
storage/myisammrg/ha_myisammrg.h:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added elements to class ha_myisammrg to support the new
open approach.
Changed empty destructor definition to a declaration.
Implemented in ha_myisammrg.cc.
Added declaration for methods attach_children() and
detach_children().
Added definition for method table_ptr() for use with
callback functions.
storage/myisammrg/myrg_close.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Added a check to avoid closing of MyISAM tables when the
child tables are not attached.
Added freeing of rec_per_key_part when the child tables
are not attached.
storage/myisammrg/myrg_extra.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Some ::extra() functions and ::reset() can be called when
children are detached.
storage/myisammrg/myrg_open.c:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE
corrupts a MERGE table
Kept old myrg_open() for MERGE use independent from MySQL.
Removed an always true condition in myrg_open().
Set children_attached for independent MERGE use in myrg_open().
Added myrg_parent_open(), myrg_attach_children(), and
myrg_detach_children() for the new MERGE table open approach.
mysql-test/r/merge-big.result:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test result
mysql-test/t/merge-big.test:
Bug#26379 - Combination of FLUSH TABLE and REPAIR TABLE corrupts a MERGE table
New test case
2007-11-15 20:25:43 +01:00
CHECKSUM TABLE t2;
Table Checksum
test.t2 3700403066
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
UNLOCK TABLES;
#
# Insert delayed
#
INSERT DELAYED INTO t4 VALUES(44);
DELETE FROM t4 WHERE c1 = 44;
INSERT DELAYED INTO t3 VALUES(33);
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
33
LOCK TABLES t3 WRITE, t2 WRITE, t4 WRITE, t1 WRITE;
INSERT DELAYED INTO t4 VALUES(444);
Got one of the listed errors
INSERT DELAYED INTO t3 VALUES(333);
Got one of the listed errors
SELECT * FROM t4 ORDER BY c1;
c1
1
2
3
4
33
UNLOCK TABLES;
DROP TABLE t1, t2, t3, t4;
#
# Recursive inclusion of merge tables in their union clauses.
#
CREATE TABLE t1 (c1 INT, INDEX(c1));
CREATE TABLE t2 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TABLE t3 (c1 INT, INDEX(c1)) ENGINE=MRG_MYISAM UNION=(t2,t1)
INSERT_METHOD=LAST;
ALTER TABLE t2 UNION=(t3,t1);
SELECT * FROM t2;
ERROR HY000: Table 't3' is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, t2, t3;
CREATE TABLE t1 (c1 INT) ENGINE= MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE= MyISAM;
CREATE TABLE t3 (c1 INT) ENGINE= MRG_MYISAM UNION= (t1, t2);
INSERT INTO t1 VALUES (1);
INSERT INTO t2 VALUES (2);
SELECT * FROM t3;
c1
1
2
TRUNCATE TABLE t1;
SELECT * FROM t3;
c1
2
DROP TABLE t1, t2, t3;
CREATE TABLE t1 (id INTEGER, grp TINYINT, id_rev INTEGER);
SET @rnd_max= 2147483647;
set @@read_buffer_size=2*1024*1024;
CREATE TABLE t2 SELECT * FROM t1;
INSERT INTO t1 (id, grp, id_rev) SELECT id, grp, id_rev FROM t2;
INSERT INTO t2 (id, grp, id_rev) SELECT id, grp, id_rev FROM t1;
INSERT INTO t1 (id, grp, id_rev) SELECT id, grp, id_rev FROM t2;
INSERT INTO t2 (id, grp, id_rev) SELECT id, grp, id_rev FROM t1;
INSERT INTO t1 (id, grp, id_rev) SELECT id, grp, id_rev FROM t2;
CREATE TABLE t3 (id INTEGER, grp TINYINT, id_rev INTEGER)
ENGINE= MRG_MYISAM UNION= (t1, t2);
SELECT COUNT(*) FROM t1;
COUNT(*)
130
SELECT COUNT(*) FROM t2;
COUNT(*)
80
SELECT COUNT(*) FROM t3;
COUNT(*)
210
SELECT COUNT(DISTINCT a1.id) FROM t3 AS a1, t3 AS a2
WHERE a1.id = a2.id GROUP BY a2.grp;
TRUNCATE TABLE t1;
SELECT COUNT(*) FROM t1;
COUNT(*)
0
SELECT COUNT(*) FROM t2;
COUNT(*)
80
SELECT COUNT(*) FROM t3;
COUNT(*)
80
DROP TABLE t1, t2, t3;
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
INSERT INTO t2 VALUES (1);
SELECT * FROM t2;
c1
1
LOCK TABLES t2 WRITE, t1 WRITE;
FLUSH TABLES;
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair status OK
CHECK TABLE t1;
Table Op Msg_type Msg_text
test.t1 check status OK
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair status OK
UNLOCK TABLES;
CHECK TABLE t1 EXTENDED;
Table Op Msg_type Msg_text
test.t1 check status OK
LOCK TABLES t2 WRITE, t1 WRITE;
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair status OK
CHECK TABLE t1;
Table Op Msg_type Msg_text
test.t1 check status OK
REPAIR TABLE t1;
Table Op Msg_type Msg_text
test.t1 repair status OK
UNLOCK TABLES;
CHECK TABLE t1 EXTENDED;
Table Op Msg_type Msg_text
test.t1 check status OK
DROP TABLE t1, t2;
CREATE TABLE t1 ( a INT ) ENGINE=MyISAM;
CREATE TABLE m1 ( a INT ) ENGINE=MRG_MYISAM UNION=(t1);
LOCK TABLES t1 WRITE, m1 WRITE;
FLUSH TABLE t1;
UNLOCK TABLES;
DROP TABLE m1, t1;
CREATE TABLE t1 ( a INT ) ENGINE=MyISAM;
CREATE TABLE m1 ( a INT ) ENGINE=MRG_MYISAM UNION=(t1);
LOCK TABLES m1 WRITE, t1 WRITE;
FLUSH TABLE t1;
UNLOCK TABLES;
DROP TABLE m1, t1;
CREATE TABLE t1 (c1 INT, c2 INT) ENGINE= MyISAM;
CREATE TABLE t2 (c1 INT, c2 INT) ENGINE= MyISAM;
CREATE TABLE t3 (c1 INT, c2 INT) ENGINE= MRG_MYISAM UNION(t1, t2);
INSERT INTO t1 VALUES (1, 1);
INSERT INTO t2 VALUES (2, 2);
SELECT * FROM t3;
c1 c2
1 1
2 2
ALTER TABLE t1 ENGINE= MEMORY;
INSERT INTO t1 VALUES (0, 0);
SELECT * FROM t3;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, t2, t3;
CREATE TABLE t1 (c1 INT, KEY(c1));
CREATE TABLE t2 (c1 INT, KEY(c1)) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=FIRST;
LOCK TABLE t1 WRITE, t2 WRITE;
FLUSH TABLES t2, t1;
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize status Table is already up to date
FLUSH TABLES t1;
UNLOCK TABLES;
FLUSH TABLES;
INSERT INTO t1 VALUES (1);
LOCK TABLE t1 WRITE, t2 WRITE;
FLUSH TABLES t2, t1;
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize status OK
FLUSH TABLES t1;
UNLOCK TABLES;
DROP TABLE t1, t2;
CREATE TABLE t1 (ID INT) ENGINE=MYISAM;
CREATE TABLE m1 (ID INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=FIRST;
INSERT INTO t1 VALUES ();
INSERT INTO m1 VALUES ();
LOCK TABLE t1 WRITE, m1 WRITE;
FLUSH TABLES m1, t1;
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize status OK
FLUSH TABLES m1, t1;
UNLOCK TABLES;
DROP TABLE t1, m1;
2008-10-09 10:55:16 +02:00
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=FIRST;
SELECT * FROM INFORMATION_SCHEMA.TABLES WHERE
TABLE_SCHEMA = 'test' and TABLE_NAME='tm1';
TABLE_CATALOG TABLE_SCHEMA TABLE_NAME TABLE_TYPE ENGINE VERSION ROW_FORMAT TABLE_ROWS AVG_ROW_LENGTH DATA_LENGTH MAX_DATA_LENGTH INDEX_LENGTH DATA_FREE AUTO_INCREMENT CREATE_TIME UPDATE_TIME CHECK_TIME TABLE_COLLATION CHECKSUM CREATE_OPTIONS TABLE_COMMENT
2010-02-20 11:07:32 +01:00
def test tm1 BASE TABLE NULL NULL NULL # # # # # # # # # # NULL # # Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2010-11-30 18:20:56 +01:00
Warnings:
Warning 1168 Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
2008-10-09 10:55:16 +02:00
DROP TABLE tm1;
2008-04-25 23:45:58 +02:00
CREATE TABLE t1(C1 INT, C2 INT, KEY C1(C1), KEY C2(C2)) ENGINE=MYISAM;
CREATE TABLE t2(C1 INT, C2 INT, KEY C1(C1), KEY C2(C2)) ENGINE=MYISAM;
CREATE TABLE t3(C1 INT, C2 INT, KEY C1(C1), KEY C2(C2)) ENGINE=MYISAM;
CREATE TABLE t4(C1 INT, C2 INT, KEY C1(C1), KEY C2(C2))
ENGINE=MRG_MYISAM UNION=(t1, t2, t3);
INSERT INTO t1 VALUES (1,1), (1,2),(1,3), (1,4);
INSERT INTO t2 VALUES (2,1), (2,2),(2,3), (2,4);
INSERT INTO t3 VALUES (3,1), (3,2),(3,3), (3,4);
EXPLAIN SELECT COUNT(*) FROM t1;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away
EXPLAIN SELECT COUNT(*) FROM t4;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away
DROP TABLE t1, t2, t3, t4;
2009-02-05 14:03:47 +01:00
CREATE TABLE t1(a INT, KEY(a));
INSERT INTO t1 VALUES(0),(1),(2),(3),(4);
ANALYZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 analyze status OK
CREATE TABLE m1(a INT, KEY(a)) ENGINE=MERGE UNION=(t1);
SELECT CARDINALITY FROM INFORMATION_SCHEMA.STATISTICS WHERE TABLE_SCHEMA='test' AND TABLE_NAME='m1';
CARDINALITY
5
SELECT CARDINALITY FROM INFORMATION_SCHEMA.STATISTICS WHERE TABLE_SCHEMA='test' AND TABLE_NAME='m1';
CARDINALITY
5
SELECT CARDINALITY FROM INFORMATION_SCHEMA.STATISTICS WHERE TABLE_SCHEMA='test' AND TABLE_NAME='m1';
CARDINALITY
5
SELECT CARDINALITY FROM INFORMATION_SCHEMA.STATISTICS WHERE TABLE_SCHEMA='test' AND TABLE_NAME='m1';
CARDINALITY
5
DROP TABLE t1, m1;
2009-02-12 12:12:07 +01:00
#
# Bug #40675 MySQL 5.1 crash with index merge algorithm and Merge tables
#
# create MYISAM table t1 and insert values into it
CREATE TABLE t1(a INT);
INSERT INTO t1 VALUES(1);
# create MYISAM table t2 and insert values into it
CREATE TABLE t2(a INT, b INT, dummy CHAR(16) DEFAULT '', KEY(a), KEY(b));
INSERT INTO t2(a,b) VALUES
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),(0,0),
(1,2);
# Create the merge table t3
CREATE TABLE t3(a INT, b INT, dummy CHAR(16) DEFAULT '', KEY(a), KEY(b))
ENGINE=MERGE UNION=(t2) INSERT_METHOD=FIRST;
# Lock tables t1 and t3 for write
LOCK TABLES t1 WRITE, t3 WRITE;
# Insert values into the merge table t3
INSERT INTO t3(a,b) VALUES(1,2);
# select from the join of t2 and t3 (The merge table)
SELECT t3.a FROM t1,t3 WHERE t3.b=2 AND t3.a=1;
a
1
1
# Unlock the tables
UNLOCK TABLES;
# drop the created tables
DROP TABLE t1, t2, t3;
2009-03-04 10:18:07 +01:00
# insert duplicate value in child table while merge table doesn't have key
create table t1 (
col1 int(10),
primary key (col1)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;
CREATE TABLE m1 (
col1 int(10) NOT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(t1);
insert into m1 (col1) values (1);
insert into m1 (col1) values (1);
ERROR 23000: Duplicate entry '' for key '*UNKNOWN*'
drop table m1, t1;
2009-07-30 12:34:41 +02:00
#
# Bug#45800 crash when replacing into a merge table and there is a duplicate
#
# Replace duplicate value in child table when merge table doesn't have key
CREATE TABLE t1 (c1 INT PRIMARY KEY) ENGINE=MyISAM;
CREATE TABLE m1 (c1 INT NOT NULL) ENGINE=MRG_MyISAM INSERT_METHOD=LAST UNION=(t1);
INSERT INTO m1 VALUES (666);
SELECT * FROM m1;
c1
666
# insert the duplicate value into the merge table
REPLACE INTO m1 VALUES (666);
SELECT * FROM m1;
c1
666
DROP TABLE m1, t1;
# Insert... on duplicate key update (with duplicate values in the table)
CREATE TABLE t1 (c1 INT PRIMARY KEY) ENGINE=MyISAM;
CREATE TABLE m1 (c1 INT NOT NULL) ENGINE=MRG_MyISAM INSERT_METHOD=LAST UNION=(t1);
INSERT INTO m1 VALUES (666);
SELECT * FROM m1;
c1
666
# insert the duplicate value into the merge table
INSERT INTO m1 VALUES (666) ON DUPLICATE KEY UPDATE c1=c1+1;
SELECT * FROM m1;
c1
667
DROP TABLE m1, t1;
# Insert duplicate value on MERGE table, where, MERGE has a key but MyISAM has more keys
CREATE TABLE t1 (c1 INT, c2 INT, UNIQUE (c1), UNIQUE (c2));
CREATE TABLE m1 (c1 INT, c2 INT, UNIQUE (c1)) ENGINE=MRG_MyISAM INSERT_METHOD=LAST UNION=(t1);
INSERT INTO m1 VALUES (1,2);
# insert the duplicate value into the merge table
INSERT INTO m1 VALUES (3,2);
ERROR 23000: Duplicate entry '' for key '*UNKNOWN*'
DROP TABLE m1,t1;
# Try to define MERGE and MyISAM with keys on different columns
CREATE TABLE t1 (c1 INT, c2 INT, UNIQUE (c1));
CREATE TABLE m1 (c1 INT, c2 INT, UNIQUE (c2)) ENGINE=MRG_MyISAM INSERT_METHOD=LAST UNION=(t1);
# Try accessing the merge table for inserts (error occurs)
INSERT INTO m1 VALUES (1,2);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
INSERT INTO m1 VALUES (1,4);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE m1,t1;
2009-04-17 07:51:51 +02:00
CREATE TABLE t1 (
col1 INT(10)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;
CREATE VIEW v1 as SELECT * FROM t1;
CREATE TABLE m1 (
col1 INT(10)
)ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(v1);
#Select should detect that the child table is a view and fail.
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP VIEW v1;
DROP TABLE m1, t1;
2009-07-10 13:34:03 +02:00
#
# Bug #45796: invalid memory reads and writes when altering merge and
# base tables
#
CREATE TABLE t1(c1 INT) ENGINE=MyISAM;
CREATE TABLE m1(c1 INT) ENGINE=MERGE UNION=(t1);
ALTER TABLE m1 ADD INDEX idx_c1(c1);
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
ALTER TABLE t1 ADD INDEX idx_c1(c1);
SELECT * FROM m1;
c1
DROP TABLE m1;
DROP TABLE t1;
2009-07-16 01:23:57 +02:00
#
# Bug45781 infinite hang/crash in "opening tables" after handler tries to
# open merge table
#
DROP TABLE IF EXISTS m1,t1;
CREATE TABLE t1(a int)engine=myisam;
CREATE TABLE t2(a int)engine=myisam;
CREATE TABLE t3(a int)engine=myisam;
CREATE TABLE t4(a int)engine=myisam;
CREATE TABLE t5(a int)engine=myisam;
CREATE TABLE t6(a int)engine=myisam;
CREATE TABLE t7(a int)engine=myisam;
CREATE TABLE m1(a int)engine=merge union=(t1,t2,t3,t4,t5,t6,t7);
SELECT 1 FROM m1;
1
HANDLER m1 OPEN;
ERROR HY000: Table storage engine for 'm1' doesn't have this option
DROP TABLE m1,t1,t2,t3,t4,t5,t6,t7;
SELECT 1 FROM m1;
ERROR 42S02: Table 'test.m1' doesn't exist
2009-08-13 21:49:28 +02:00
#
# Bug #46614: Assertion in show_create_trigger()
#
CREATE TABLE t1(a int);
CREATE TABLE t2(a int);
CREATE TABLE t3(a int) ENGINE = MERGE UNION(t1, t2);
CREATE TRIGGER tr1 AFTER INSERT ON t3 FOR EACH ROW CALL foo();
SHOW CREATE TRIGGER tr1;
Trigger sql_mode SQL Original Statement character_set_client collation_connection Database Collation
tr1 CREATE DEFINER=`root`@`localhost` TRIGGER tr1 AFTER INSERT ON t3 FOR EACH ROW CALL foo() latin1 latin1_swedish_ci latin1_swedish_ci
DROP TRIGGER tr1;
DROP TABLE t1, t2, t3;
2010-03-03 11:49:03 +01:00
#
# BUG#48265 - MRG_MYISAM problem (works in 5.0.85, does't work in 5.1.40)
#
CREATE DATABASE `test/1`;
CREATE TABLE `test/1`.`t/1`(a INT);
CREATE TABLE m1(a INT) ENGINE=MERGE UNION=(`test/1`.`t/1`);
SELECT * FROM m1;
a
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`test/1`.`t/1`)
DROP TABLE m1;
CREATE TABLE `test/1`.m1(a INT) ENGINE=MERGE UNION=(`test/1`.`t/1`);
SELECT * FROM `test/1`.m1;
a
SHOW CREATE TABLE `test/1`.m1;
Table Create Table
m1 CREATE TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t/1`)
DROP TABLE `test/1`.m1;
DROP TABLE `test/1`.`t/1`;
CREATE TEMPORARY TABLE `test/1`.`t/1`(a INT);
CREATE TEMPORARY TABLE m1(a INT) ENGINE=MERGE UNION=(`test/1`.`t/1`);
SELECT * FROM m1;
a
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`test/1`.`t/1`)
DROP TABLE m1;
CREATE TEMPORARY TABLE `test/1`.m1(a INT) ENGINE=MERGE UNION=(`test/1`.`t/1`);
SELECT * FROM `test/1`.m1;
a
SHOW CREATE TABLE `test/1`.m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t/1`)
DROP TABLE `test/1`.m1;
DROP TABLE `test/1`.`t/1`;
DROP DATABASE `test/1`;
CREATE TABLE `t@1`(a INT);
SELECT * FROM m1;
a
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`t@1`)
DROP TABLE `t@1`;
CREATE DATABASE `test@1`;
CREATE TABLE `test@1`.`t@1`(a INT);
FLUSH TABLE m1;
SELECT * FROM m1;
a
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TABLE `m1` (
`a` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 UNION=(`test@1`.`t@1`)
DROP TABLE m1;
DROP TABLE `test@1`.`t@1`;
DROP DATABASE `test@1`;
2010-03-19 07:01:02 +01:00
#
# Bug#51494c rash with join, explain and 'sounds like' operator
#
CREATE TABLE t1 (a INT) ENGINE=MYISAM;
INSERT INTO t1 VALUES(1);
CREATE TABLE t2 (b INT NOT NULL,c INT,d INT,e BLOB NOT NULL,
KEY idx0 (d, c)) ENGINE=MERGE;
EXPLAIN SELECT * FROM t1 NATURAL RIGHT JOIN
t2 WHERE b SOUNDS LIKE e AND d = 1;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables
DROP TABLE t2, t1;
Bug #56494 Segfault in upgrade_shared_lock_to_exclusive() for
REPAIR of merge table
Bug #56422 CHECK TABLE run when the table is locked reports
corruption along with timeout
The crash happened if a table maintenance statement (ANALYZE TABLE,
REPAIR TABLE, etc.) was executed on a MERGE table and opening and
locking a child table failed. This could for example happen if a child
table did not exist or if a lock timeout happened while waiting for
a conflicting metadata lock to disappear.
Since opening and locking the MERGE table and its children failed,
the tables would be closed and the metadata locks released.
However, TABLE_LIST::table for the MERGE table would still be set,
with its value invalid since the tables had been closed.
This caused the table maintenance statement to try to continue
and upgrade the metadata lock on the MERGE table. But since the lock
already had been released, this caused a segfault.
This patch fixes the problem by setting TABLE_LIST::table to NULL
if open_and_lock_tables() fails. This prevents maintenance
statements from continuing and trying to upgrade the metadata lock.
The patch includes a 5.5 version of the fix for
Bug #46339 crash on REPAIR TABLE merge table USE_FRM.
This bug caused REPAIR TABLE ... USE_FRM to give an assert
when used on merge tables.
The patch also enables the CHECK TABLE statement for log tables.
Before, CHECK TABLE for log tables gave ER_CANT_LOCK_LOG_TABLE,
yet still counted the statement as successfully executed.
With the changes to table maintenance statement error handling
in this patch, CHECK TABLE would no longer be considered as
successful in this case. This would have caused upgrade scripts
to mistakenly think that the general and slow logs are corrupted
and have to be repaired. Enabling CHECK TABLES for log tables
prevents this from happening.
Finally, the patch changes the error message from "Corrupt" to
"Operation failed" for a number of issues not related to table
corruption. For example "Lock wait timeout exceeded" and
"Deadlock found trying to get lock".
Test cases added to merge.test and check.test.
2010-09-22 10:15:41 +02:00
#
# Bug#46339 - crash on REPAIR TABLE merge table USE_FRM
#
DROP TABLE IF EXISTS m1, t1;
CREATE TABLE t1 (c1 INT) ENGINE=MYISAM;
CREATE TABLE m1 (c1 INT) ENGINE=MRG_MyISAM UNION=(t1) INSERT_METHOD=LAST;
LOCK TABLE m1 READ;
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
test.m1 repair Error Table 'm1' was locked with a READ lock and can't be updated
test.m1 repair status Operation failed
UNLOCK TABLES;
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
test.m1 repair note The storage engine for the table doesn't support repair
DROP TABLE m1,t1;
CREATE TABLE m1 (f1 BIGINT) ENGINE=MRG_MyISAM UNION(t1);
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
test.m1 repair Warning Can't open table
test.m1 repair error Corrupt
CREATE TABLE t1 (f1 BIGINT) ENGINE = MyISAM;
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
test.m1 repair note The storage engine for the table doesn't support repair
REPAIR TABLE m1;
Table Op Msg_type Msg_text
test.m1 repair note The storage engine for the table doesn't support repair
DROP TABLE m1, t1;
CREATE TEMPORARY TABLE m1 (f1 BIGINT) ENGINE=MRG_MyISAM UNION(t1);
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
test.m1 repair Error Table 'test.m1' doesn't exist
test.m1 repair error Corrupt
CREATE TEMPORARY TABLE t1 (f1 BIGINT) ENGINE=MyISAM;
REPAIR TABLE m1 USE_FRM;
Table Op Msg_type Msg_text
m1 repair error Cannot repair temporary table from .frm file
REPAIR TABLE m1;
Table Op Msg_type Msg_text
test.m1 repair note The storage engine for the table doesn't support repair
DROP TABLE m1, t1;
2013-11-29 20:21:05 +01:00
create temporary table t1_temp(i int);
create temporary table tm_temp_temp (i int) engine=merge union=(t1_temp) insert_method=last;
alter table tm_temp_temp insert_method=first;
check table tm_temp_temp;
Table Op Msg_type Msg_text
test.tm_temp_temp check status OK
drop temporary table t1_temp, tm_temp_temp;
2008-04-25 23:45:58 +02:00
End of 5.1 tests
2013-03-21 08:06:27 +01:00
#
# MDEV-4277: Crash inside mi_killed_in_mariadb() with myisammrg
#
create table t1 (a int);
insert into t1 values (0),(1),(2),(3),(4),(5),(6),(7),(8),(9);
create table t2 ( i int ) engine=myisam;
insert into t2 values (1),(2);
create table t3 (a int, b int, filler char(100), key(a), key(b)) engine=myisam;
create table t4 like t3;
insert into t3
select A.a+10*B.a+100*C.a,
A.a+10*B.a+100*C.a,
'filler-data-FILLER-DATA-qqq'
from t1 A, t1 B, t1 C where C.a < 5;
insert into t4
select A.a+10*B.a+100*C.a,
A.a+10*B.a+100*C.a,
'filler-data-FILLER-DATA-qqq'
from t1 A, t1 B, t1 C where C.a >= 5;
create table t5 like t3;
alter table t5 engine=merge;
alter table t5 union(t3, t4);
update t5 set b=999, a=999 where b>950;
explain
select * from t2, t5 where t5.a=999 and t5.b=999;
id select_type table type possible_keys key key_len ref rows Extra
1 SIMPLE t2 ALL NULL NULL NULL NULL 2
1 SIMPLE t5 index_merge a,b a,b 5,5 NULL 6 Using intersect(a,b); Using where; Using join buffer (flat, BNL join)
select * from t2, t5 where t5.a=999 and t5.b=999;
i a b filler
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
1 999 999 filler-data-FILLER-DATA-qqq
2 999 999 filler-data-FILLER-DATA-qqq
drop table t5;
drop table t1,t2,t3,t4;
2013-03-27 10:03:28 +01:00
End of 5.3 tests
CREATE TABLE t1(a INT, KEY(a)) ENGINE=merge;
SELECT MAX(a) FROM t1;
MAX(a)
NULL
DROP TABLE t1;
2009-10-13 21:04:58 +02:00
#
# An additional test case for Bug#27430 Crash in subquery code
# when in PS and table DDL changed after PREPARE
#
# Test merge table with too many merge children.
#
drop table if exists t_parent;
set @save_table_definition_cache=@@global.table_definition_cache;
#
# Set @@global.table_definition_cache to minimum
#
set @@global.table_definition_cache=400;
set @a=null;
#
# Create 400 merge children
#
set @a=concat("create table t_parent (a int) union(", @a,
") insert_method=first engine=mrg_myisam");
prepare stmt from @a;
execute stmt;
prepare stmt from "select * from t_parent";
execute stmt;
ERROR HY000: Prepared statement needs to be re-prepared
execute stmt;
ERROR HY000: Prepared statement needs to be re-prepared
execute stmt;
ERROR HY000: Prepared statement needs to be re-prepared
deallocate prepare stmt;
#
# Create merge parent
#
#
# Cleanup
#
drop table t_parent;
set @@global.table_definition_cache=@save_table_definition_cache;
2009-12-03 00:09:22 +01:00
DROP DATABASE IF EXISTS mysql_test1;
CREATE DATABASE mysql_test1;
CREATE TABLE t1 ... DATA DIRECTORY=... INDEX DIRECTORY=...
CREATE TABLE mysql_test1.t2 ... DATA DIRECTORY=... INDEX DIRECTORY=...
CREATE TABLE m1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,mysql_test1.t2)
INSERT_METHOD=LAST;
INSERT INTO t1 VALUES (1);
INSERT INTO mysql_test1.t2 VALUES (2);
SELECT * FROM m1;
c1
1
2
DROP TABLE t1, mysql_test1.t2, m1;
DROP DATABASE mysql_test1;
CREATE TABLE t1 (c1 INT);
CREATE TABLE t2 (c1 INT);
INSERT INTO t1 (c1) VALUES (1);
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2) INSERT_METHOD=FIRST;
CREATE TABLE t3 (c1 INT);
INSERT INTO t3 (c1) VALUES (1);
CREATE FUNCTION f1() RETURNS INT RETURN (SELECT MAX(c1) FROM t3);
CREATE VIEW v1 AS SELECT foo.c1 c1, f1() c2, bar.c1 c3, f1() c4
FROM tm1 foo, tm1 bar, t3;
SELECT * FROM v1;
c1 c2 c3 c4
1 1 1 1
DROP FUNCTION f1;
DROP VIEW v1;
DROP TABLE tm1, t1, t2, t3;
CREATE TEMPORARY TABLE t1 (c1 INT);
CREATE TEMPORARY TABLE t2 (c1 INT);
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2)
INSERT_METHOD=FIRST;
CREATE FUNCTION f1() RETURNS INT RETURN (SELECT MAX(c1) FROM tm1);
INSERT INTO tm1 (c1) VALUES (1);
SELECT f1() FROM (SELECT 1) AS c1;
f1()
1
DROP FUNCTION f1;
DROP TABLE tm1, t1, t2;
CREATE FUNCTION f1() RETURNS INT
BEGIN
CREATE TEMPORARY TABLE t1 (c1 INT);
CREATE TEMPORARY TABLE t2 (c1 INT);
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2);
INSERT INTO t1 (c1) VALUES (1);
RETURN (SELECT MAX(c1) FROM tm1);
END|
SELECT f1() FROM (SELECT 1 UNION SELECT 1) c1;
f1()
1
DROP FUNCTION f1;
DROP TABLE tm1, t1, t2;
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
INSERT INTO tm1 VALUES (1);
SELECT * FROM tm1;
c1
1
DROP TABLE tm1, t1;
CREATE FUNCTION f1() RETURNS INT
BEGIN
INSERT INTO tm1 VALUES (1);
RETURN (SELECT MAX(c1) FROM tm1);
END|
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
SELECT f1();
f1()
1
DROP FUNCTION f1;
DROP TABLE tm1, t1;
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
LOCK TABLE tm1 WRITE;
INSERT INTO tm1 VALUES (1);
SELECT * FROM tm1;
c1
1
UNLOCK TABLES;
DROP TABLE tm1, t1;
CREATE FUNCTION f1() RETURNS INT
BEGIN
INSERT INTO tm1 VALUES (1);
RETURN (SELECT MAX(c1) FROM tm1);
END|
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
LOCK TABLE tm1 WRITE;
SELECT f1();
f1()
1
UNLOCK TABLES;
DROP FUNCTION f1;
DROP TABLE tm1, t1;
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1) INSERT_METHOD=LAST;
CREATE TRIGGER t2_ai AFTER INSERT ON t2
FOR EACH ROW INSERT INTO tm1 VALUES(11);
LOCK TABLE t2 WRITE;
INSERT INTO t2 VALUES (2);
SELECT * FROM tm1;
c1
11
SELECT * FROM t2;
c1
2
UNLOCK TABLES;
DROP TRIGGER t2_ai;
DROP TABLE tm1, t1, t2;
CREATE TEMPORARY TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
INSERT INTO tm1 VALUES (1);
SELECT * FROM tm1;
c1
1
DROP TABLE tm1, t1;
CREATE FUNCTION f1() RETURNS INT
BEGIN
INSERT INTO tm1 VALUES (1);
RETURN (SELECT MAX(c1) FROM tm1);
END|
CREATE TEMPORARY TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
SELECT f1();
f1()
1
DROP FUNCTION f1;
DROP TABLE tm1, t1;
CREATE TEMPORARY TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TABLE t9 (c1 INT) ENGINE=MyISAM;
LOCK TABLE t9 WRITE;
INSERT INTO tm1 VALUES (1);
SELECT * FROM tm1;
c1
1
UNLOCK TABLES;
DROP TABLE tm1, t1, t9;
CREATE FUNCTION f1() RETURNS INT
BEGIN
INSERT INTO tm1 VALUES (1);
RETURN (SELECT MAX(c1) FROM tm1);
END|
CREATE TEMPORARY TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TABLE t9 (c1 INT) ENGINE=MyISAM;
LOCK TABLE t9 WRITE;
SELECT f1();
f1()
1
UNLOCK TABLES;
DROP FUNCTION f1;
DROP TABLE tm1, t1, t9;
CREATE TEMPORARY TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TABLE t2 (c1 INT) ENGINE=MyISAM;
CREATE TRIGGER t2_ai AFTER INSERT ON t2
FOR EACH ROW INSERT INTO tm1 VALUES(11);
LOCK TABLE t2 WRITE;
INSERT INTO t2 VALUES (2);
SELECT * FROM tm1;
c1
11
SELECT * FROM t2;
c1
2
UNLOCK TABLES;
DROP TRIGGER t2_ai;
DROP TABLE tm1, t1, t2;
#
A fix and a test case for Bug#47648 "main.merge fails sporadically".
If a prepared statement used both a MyISAMMRG table and a stored
function or trigger, execution could fail with "No such table"
error or crash.
The error would come from a failure of the MyISAMMRG engine
to meet the expectations of the prelocking algorithm,
in particular maintain lex->query_tables_own_last pointer
in sync with lex->query_tables_last pointer/the contents
of lex->query_tables. When adding merge children, the merge
engine would extend the table list. Then, when adding
prelocked tables, the prelocking algorithm would use a pointer
to the last merge child to assign to lex->query_tables_own_last.
Then, when merge children were removed at the end of
open_tables(), lex->query_tables_own_last
was not updated, and kept pointing
to a removed merge child.
The fix ensures that query_tables_own_last is always in
sync with lex->query_tables_last.
This is a regression introduced by WL#4144 and present only
in next-4284 tree and 6.0.
mysql-test/r/merge.result:
Update results (Bug#47648).
mysql-test/t/merge.test:
Add a test case for Bug#47648.
Update the result file to reflect a fix of another bug
in MyISAMMRG code: not maintaining lex->query_tables_own_last
allowed a stored function or trigger to modify a merge table
which was already updated by the main statement.
It is not allowed for other storage engines, and should
not be allowed for MyISAMMRG.
storage/myisammrg/ha_myisammrg.cc:
When adding children to the list of tables to open,
make sure that we properly set lex->query_tables_own_last.
When removing the children, update lex->query_tables_own_last
if necessary.
2010-02-15 17:35:53 +01:00
# Don't allow an update of a MERGE child in a trigger
# if the table's already being modified by the main
# statement.
2009-12-03 00:09:22 +01:00
#
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TRIGGER tm1_ai AFTER INSERT ON tm1
FOR EACH ROW INSERT INTO t1 VALUES(11);
LOCK TABLE tm1 WRITE, t1 WRITE;
INSERT INTO tm1 VALUES (1);
A fix and a test case for Bug#47648 "main.merge fails sporadically".
If a prepared statement used both a MyISAMMRG table and a stored
function or trigger, execution could fail with "No such table"
error or crash.
The error would come from a failure of the MyISAMMRG engine
to meet the expectations of the prelocking algorithm,
in particular maintain lex->query_tables_own_last pointer
in sync with lex->query_tables_last pointer/the contents
of lex->query_tables. When adding merge children, the merge
engine would extend the table list. Then, when adding
prelocked tables, the prelocking algorithm would use a pointer
to the last merge child to assign to lex->query_tables_own_last.
Then, when merge children were removed at the end of
open_tables(), lex->query_tables_own_last
was not updated, and kept pointing
to a removed merge child.
The fix ensures that query_tables_own_last is always in
sync with lex->query_tables_last.
This is a regression introduced by WL#4144 and present only
in next-4284 tree and 6.0.
mysql-test/r/merge.result:
Update results (Bug#47648).
mysql-test/t/merge.test:
Add a test case for Bug#47648.
Update the result file to reflect a fix of another bug
in MyISAMMRG code: not maintaining lex->query_tables_own_last
allowed a stored function or trigger to modify a merge table
which was already updated by the main statement.
It is not allowed for other storage engines, and should
not be allowed for MyISAMMRG.
storage/myisammrg/ha_myisammrg.cc:
When adding children to the list of tables to open,
make sure that we properly set lex->query_tables_own_last.
When removing the children, update lex->query_tables_own_last
if necessary.
2010-02-15 17:35:53 +01:00
ERROR HY000: Can't update table 't1' in stored function/trigger because it is already used by statement which invoked this stored function/trigger.
SELECT * FROM tm1;
c1
1
UNLOCK TABLES;
LOCK TABLE t1 WRITE, tm1 WRITE;
INSERT INTO tm1 VALUES (1);
ERROR HY000: Can't update table 't1' in stored function/trigger because it is already used by statement which invoked this stored function/trigger.
SELECT * FROM tm1;
c1
1
1
UNLOCK TABLES;
DROP TRIGGER tm1_ai;
DROP TABLE tm1, t1;
#
# Don't select MERGE child when trying to get a prelocked table.
#
# Due to a limitation demonstrated by the previous test
# we can no longer use a write-locked prelocked table.
# The test is kept for historical purposes.
#
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1)
INSERT_METHOD=LAST;
CREATE TRIGGER tm1_ai AFTER INSERT ON tm1
FOR EACH ROW SELECT max(c1) FROM t1 INTO @var;
LOCK TABLE tm1 WRITE, t1 WRITE;
INSERT INTO tm1 VALUES (1);
2009-12-03 00:09:22 +01:00
SELECT * FROM tm1;
c1
1
UNLOCK TABLES;
LOCK TABLE t1 WRITE, tm1 WRITE;
INSERT INTO tm1 VALUES (1);
SELECT * FROM tm1;
c1
1
1
UNLOCK TABLES;
DROP TRIGGER tm1_ai;
DROP TABLE tm1, t1;
CREATE TABLE t1 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t2 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t3 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t4 (c1 INT) ENGINE=MyISAM;
CREATE TABLE t5 (c1 INT) ENGINE=MyISAM;
CREATE TABLE tm1 (c1 INT) ENGINE=MRG_MYISAM UNION=(t1,t2,t3,t4,t5)
INSERT_METHOD=LAST;
CREATE TRIGGER t2_au AFTER UPDATE ON t2
A fix and a test case for Bug#47648 "main.merge fails sporadically".
If a prepared statement used both a MyISAMMRG table and a stored
function or trigger, execution could fail with "No such table"
error or crash.
The error would come from a failure of the MyISAMMRG engine
to meet the expectations of the prelocking algorithm,
in particular maintain lex->query_tables_own_last pointer
in sync with lex->query_tables_last pointer/the contents
of lex->query_tables. When adding merge children, the merge
engine would extend the table list. Then, when adding
prelocked tables, the prelocking algorithm would use a pointer
to the last merge child to assign to lex->query_tables_own_last.
Then, when merge children were removed at the end of
open_tables(), lex->query_tables_own_last
was not updated, and kept pointing
to a removed merge child.
The fix ensures that query_tables_own_last is always in
sync with lex->query_tables_last.
This is a regression introduced by WL#4144 and present only
in next-4284 tree and 6.0.
mysql-test/r/merge.result:
Update results (Bug#47648).
mysql-test/t/merge.test:
Add a test case for Bug#47648.
Update the result file to reflect a fix of another bug
in MyISAMMRG code: not maintaining lex->query_tables_own_last
allowed a stored function or trigger to modify a merge table
which was already updated by the main statement.
It is not allowed for other storage engines, and should
not be allowed for MyISAMMRG.
storage/myisammrg/ha_myisammrg.cc:
When adding children to the list of tables to open,
make sure that we properly set lex->query_tables_own_last.
When removing the children, update lex->query_tables_own_last
if necessary.
2010-02-15 17:35:53 +01:00
FOR EACH ROW SELECT MAX(c1) FROM t1 INTO @var;
2009-12-03 00:09:22 +01:00
CREATE FUNCTION f1() RETURNS INT
RETURN (SELECT MAX(c1) FROM t4);
LOCK TABLE tm1 WRITE, t1 WRITE, t2 WRITE, t3 WRITE, t4 WRITE, t5 WRITE;
INSERT INTO t1 VALUES(1);
INSERT INTO t2 VALUES(2);
INSERT INTO t3 VALUES(3);
INSERT INTO t4 VALUES(4);
INSERT INTO t5 VALUES(5);
UPDATE t2, tm1 SET t2.c1=f1();
FLUSH TABLES;
FLUSH TABLES;
UNLOCK TABLES;
SELECT * FROM tm1;
c1
1
4
3
4
5
DROP TRIGGER t2_au;
DROP FUNCTION f1;
DROP TABLE tm1, t1, t2, t3, t4, t5;
2009-12-09 14:41:56 +01:00
#
2010-06-25 17:32:59 +02:00
# Bug#47633 - assert in ha_myisammrg::info during OPTIMIZE
#
CREATE TEMPORARY TABLE t1 (c1 INT);
ALTER TABLE t1 ENGINE=MERGE UNION(t_not_exists,t1);
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize Error Table 'test.t_not_exists' doesn't exist
test.t1 optimize Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
Bug #56494 Segfault in upgrade_shared_lock_to_exclusive() for
REPAIR of merge table
Bug #56422 CHECK TABLE run when the table is locked reports
corruption along with timeout
The crash happened if a table maintenance statement (ANALYZE TABLE,
REPAIR TABLE, etc.) was executed on a MERGE table and opening and
locking a child table failed. This could for example happen if a child
table did not exist or if a lock timeout happened while waiting for
a conflicting metadata lock to disappear.
Since opening and locking the MERGE table and its children failed,
the tables would be closed and the metadata locks released.
However, TABLE_LIST::table for the MERGE table would still be set,
with its value invalid since the tables had been closed.
This caused the table maintenance statement to try to continue
and upgrade the metadata lock on the MERGE table. But since the lock
already had been released, this caused a segfault.
This patch fixes the problem by setting TABLE_LIST::table to NULL
if open_and_lock_tables() fails. This prevents maintenance
statements from continuing and trying to upgrade the metadata lock.
The patch includes a 5.5 version of the fix for
Bug #46339 crash on REPAIR TABLE merge table USE_FRM.
This bug caused REPAIR TABLE ... USE_FRM to give an assert
when used on merge tables.
The patch also enables the CHECK TABLE statement for log tables.
Before, CHECK TABLE for log tables gave ER_CANT_LOCK_LOG_TABLE,
yet still counted the statement as successfully executed.
With the changes to table maintenance statement error handling
in this patch, CHECK TABLE would no longer be considered as
successful in this case. This would have caused upgrade scripts
to mistakenly think that the general and slow logs are corrupted
and have to be repaired. Enabling CHECK TABLES for log tables
prevents this from happening.
Finally, the patch changes the error message from "Corrupt" to
"Operation failed" for a number of issues not related to table
corruption. For example "Lock wait timeout exceeded" and
"Deadlock found trying to get lock".
Test cases added to merge.test and check.test.
2010-09-22 10:15:41 +02:00
test.t1 optimize error Corrupt
2010-06-25 17:32:59 +02:00
DROP TABLE t1;
#
2010-07-02 18:07:57 +02:00
# Bug#36171 - CREATE TEMPORARY TABLE and MERGE engine
# More tests with TEMPORARY MERGE table and permanent children.
# First without locked tables.
#
DROP TABLE IF EXISTS t1, t2, t3, t4, m1, m2;
#
CREATE TABLE t1 (c1 INT, c2 INT) ENGINE=MyISAM;
CREATE TABLE t2 (c1 INT, c2 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE m1 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
SHOW CREATE TABLE t1;
Table Create Table
t1 CREATE TABLE `t1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m1;
c1 c2
INSERT INTO t1 VALUES (111, 121);
INSERT INTO m1 VALUES (211, 221);
SELECT * FROM m1;
c1 c2
111 121
211 221
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
#
ALTER TABLE m1 RENAME m2;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m2;
c1 c2
111 121
211 221
#
CREATE TEMPORARY TABLE m1 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
ALTER TABLE m2 RENAME m1;
ERROR 42S01: Table 'm1' already exists
DROP TABLE m1;
ALTER TABLE m2 RENAME m1;
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m1;
c1 c2
111 121
211 221
#
ALTER TABLE m1 ADD COLUMN c3 INT;
INSERT INTO m1 VALUES (212, 222, 232);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
ALTER TABLE t1 ADD COLUMN c3 INT;
ALTER TABLE t2 ADD COLUMN c3 INT;
INSERT INTO m1 VALUES (212, 222, 232);
SELECT * FROM m1;
c1 c2 c3
111 121 NULL
211 221 NULL
212 222 232
#
ALTER TABLE m1 DROP COLUMN c3;
INSERT INTO m1 VALUES (213, 223);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
ALTER TABLE t1 DROP COLUMN c3;
ALTER TABLE t2 DROP COLUMN c3;
INSERT INTO m1 VALUES (213, 223);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
#
CREATE TABLE t3 (c1 INT, c2 INT) ENGINE=MyISAM;
ALTER TABLE m1 UNION=(t1,t2,t3);
INSERT INTO m1 VALUES (311, 321);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
SELECT * FROM t3;
c1 c2
311 321
#
CREATE TEMPORARY TABLE t4 (c1 INT, c2 INT) ENGINE=MyISAM;
ALTER TABLE m1 UNION=(t1,t2,t3,t4);
INSERT INTO m1 VALUES (411, 421);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
411 421
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
SELECT * FROM t3;
c1 c2
311 321
SELECT * FROM t4;
c1 c2
411 421
#
ALTER TABLE m1 ENGINE=MyISAM;
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
INSERT INTO m1 VALUES (511, 521);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
411 421
511 521
#
ALTER TABLE m1 ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
#
CREATE TEMPORARY TABLE t1 (c1 INT, c2 INT) ENGINE=MyISAM;
INSERT INTO t1 VALUES (611, 621);
SELECT * FROM m1;
c1 c2
611 621
211 221
212 222
213 223
DROP TABLE t1;
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
#
#
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
#
CREATE TABLE m2 SELECT * FROM m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TEMPORARY TABLE m2 SELECT * FROM m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST;
SELECT * FROM m2;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE m2;
#
CREATE TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST SELECT * FROM m1;
ERROR HY000: 'test.m2' is not BASE TABLE
#
CREATE TEMPORARY TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST SELECT * FROM m1;
ERROR HY000: 'test.m2' is not BASE TABLE
#
CREATE TABLE m2 LIKE m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
INSERT INTO m2 SELECT * FROM m1;
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TEMPORARY TABLE m2 LIKE m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
INSERT INTO m2 SELECT * FROM m1;
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TEMPORARY TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST;
INSERT INTO m2 SELECT * FROM m1;
SELECT * FROM m2;
c1 c2
311 321
411 421
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
#
#
LOCK TABLE m1 WRITE, m2 WRITE;
SELECT * FROM m1,m2 WHERE m1.c1=m2.c1;
c1 c2 c1 c2
111 121 111 121
111 121 111 121
111 121 111 121
111 121 111 121
211 221 211 221
211 221 211 221
211 221 211 221
211 221 211 221
212 222 212 222
212 222 212 222
212 222 212 222
212 222 212 222
213 223 213 223
213 223 213 223
213 223 213 223
213 223 213 223
111 121 111 121
111 121 111 121
111 121 111 121
111 121 111 121
211 221 211 221
211 221 211 221
211 221 211 221
211 221 211 221
212 222 212 222
212 222 212 222
212 222 212 222
212 222 212 222
213 223 213 223
213 223 213 223
213 223 213 223
213 223 213 223
111 121 111 121
111 121 111 121
111 121 111 121
111 121 111 121
211 221 211 221
211 221 211 221
211 221 211 221
211 221 211 221
212 222 212 222
212 222 212 222
212 222 212 222
212 222 212 222
213 223 213 223
213 223 213 223
213 223 213 223
213 223 213 223
111 121 111 121
111 121 111 121
111 121 111 121
111 121 111 121
211 221 211 221
211 221 211 221
211 221 211 221
211 221 211 221
212 222 212 222
212 222 212 222
212 222 212 222
212 222 212 222
213 223 213 223
213 223 213 223
213 223 213 223
213 223 213 223
UNLOCK TABLES;
DROP TABLE t1, t2, t3, t4, m1, m2;
#
# Bug#36171 - CREATE TEMPORARY TABLE and MERGE engine
# More tests with TEMPORARY MERGE table and permanent children.
# (continued) Now the same with locked table.
#
CREATE TABLE t1 (c1 INT, c2 INT) ENGINE=MyISAM;
CREATE TABLE t2 (c1 INT, c2 INT) ENGINE=MyISAM;
CREATE TEMPORARY TABLE m1 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
SHOW CREATE TABLE t1;
Table Create Table
t1 CREATE TABLE `t1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m1;
c1 c2
INSERT INTO t1 VALUES (111, 121);
INSERT INTO m1 VALUES (211, 221);
SELECT * FROM m1;
c1 c2
111 121
211 221
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
#
LOCK TABLE m1 WRITE, t1 WRITE, t2 WRITE;
#
ALTER TABLE m1 RENAME m2;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m2;
c1 c2
111 121
211 221
#
CREATE TEMPORARY TABLE m1 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
ALTER TABLE m2 RENAME m1;
ERROR 42S01: Table 'm1' already exists
DROP TABLE m1;
ALTER TABLE m2 RENAME m1;
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m1;
c1 c2
111 121
211 221
#
ALTER TABLE m1 ADD COLUMN c3 INT;
INSERT INTO m1 VALUES (212, 222, 232);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
ALTER TABLE t1 ADD COLUMN c3 INT;
ALTER TABLE t2 ADD COLUMN c3 INT;
INSERT INTO m1 VALUES (212, 222, 232);
SELECT * FROM m1;
c1 c2 c3
111 121 NULL
211 221 NULL
212 222 232
#
ALTER TABLE m1 DROP COLUMN c3;
INSERT INTO m1 VALUES (213, 223);
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
SELECT * FROM m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
ALTER TABLE t1 DROP COLUMN c3;
ALTER TABLE t2 DROP COLUMN c3;
INSERT INTO m1 VALUES (213, 223);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
#
UNLOCK TABLES;
CREATE TABLE t3 (c1 INT, c2 INT) ENGINE=MyISAM;
ALTER TABLE m1 UNION=(t1,t2,t3);
LOCK TABLE m1 WRITE;
INSERT INTO m1 VALUES (311, 321);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
SELECT * FROM t3;
c1 c2
311 321
#
CREATE TEMPORARY TABLE t4 (c1 INT, c2 INT) ENGINE=MyISAM;
ALTER TABLE m1 UNION=(t1,t2,t3,t4);
INSERT INTO m1 VALUES (411, 421);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
411 421
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
SELECT * FROM t3;
c1 c2
311 321
SELECT * FROM t4;
c1 c2
411 421
#
ALTER TABLE m1 ENGINE=MyISAM;
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
INSERT INTO m1 VALUES (511, 521);
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
311 321
411 421
511 521
#
ALTER TABLE m1 ENGINE=MRG_MyISAM UNION=(t1,t2)
INSERT_METHOD=LAST;
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
SELECT * FROM t1;
c1 c2
111 121
SELECT * FROM t2;
c1 c2
211 221
212 222
213 223
#
CREATE TEMPORARY TABLE t1 (c1 INT, c2 INT) ENGINE=MyISAM;
INSERT INTO t1 VALUES (611, 621);
SELECT * FROM m1;
c1 c2
611 621
211 221
212 222
213 223
DROP TABLE t1;
SELECT * FROM m1;
c1 c2
111 121
211 221
212 222
213 223
#
#
SHOW CREATE TABLE m1;
Table Create Table
m1 CREATE TEMPORARY TABLE `m1` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
CREATE TABLE m2 SELECT * FROM m1;
ERROR HY000: Table 'm2' was not locked with LOCK TABLES
#
CREATE TEMPORARY TABLE m2 SELECT * FROM m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TEMPORARY TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST;
SELECT * FROM m2;
c1 c2
311 321
411 421
LOCK TABLE m1 WRITE, m2 WRITE;
UNLOCK TABLES;
DROP TABLE m2;
LOCK TABLE m1 WRITE;
#
# ER_TABLE_NOT_LOCKED is returned in ps-protocol
CREATE TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST SELECT * FROM m1;
Got one of the listed errors
#
CREATE TEMPORARY TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST SELECT * FROM m1;
ERROR HY000: 'test.m2' is not BASE TABLE
#
CREATE TEMPORARY TABLE m2 LIKE m1;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
LOCK TABLE m1 WRITE, m2 WRITE;
SHOW CREATE TABLE m2;
Table Create Table
m2 CREATE TEMPORARY TABLE `m2` (
`c1` int(11) DEFAULT NULL,
`c2` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=LAST UNION=(`t1`,`t2`)
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
INSERT INTO m2 SELECT * FROM m1;
SELECT * FROM m2;
c1 c2
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
DROP TABLE m2;
#
CREATE TEMPORARY TABLE m2 (c1 INT, c2 INT) ENGINE=MRG_MyISAM UNION=(t3,t4)
INSERT_METHOD=LAST;
LOCK TABLE m1 WRITE, m2 WRITE;
INSERT INTO m2 SELECT * FROM m1;
SELECT * FROM m2;
c1 c2
311 321
411 421
111 121
211 221
212 222
213 223
111 121
211 221
212 222
213 223
#
UNLOCK TABLES;
DROP TABLE t1, t2, t3, t4, m1, m2;
#
2009-12-09 14:41:56 +01:00
# Bug47098 assert in MDL_context::destroy on HANDLER
# <damaged merge table> OPEN
#
# Test that merge tables are closed correctly when opened using
# HANDLER ... OPEN.
# The general case.
DROP TABLE IF EXISTS t1, t2, t3;
# Connection con1.
CREATE TABLE t1 (c1 int);
CREATE TABLE t2 (c1 int);
CREATE TABLE t3 (c1 int) ENGINE = MERGE UNION (t1,t2);
START TRANSACTION;
HANDLER t3 OPEN;
ERROR HY000: Table storage engine for 't3' doesn't have this option
DROP TABLE t1, t2, t3;
# Connection default.
# Disconnecting con1, all mdl_tickets must have been released.
# The bug-specific case.
# Connection con1.
CREATE TABLE t1 (c1 int);
CREATE TABLE t2 (c1 int);
CREATE TABLE t3 (c1 int) ENGINE = MERGE UNION (t1,t2);
DROP TABLE t2;
START TRANSACTION;
HANDLER t3 OPEN;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
DROP TABLE t1, t3;
# Connection default.
# Disconnecting con1, all mdl_tickets must have been released.
A fix and a test case for Bug#47648 "main.merge fails sporadically".
If a prepared statement used both a MyISAMMRG table and a stored
function or trigger, execution could fail with "No such table"
error or crash.
The error would come from a failure of the MyISAMMRG engine
to meet the expectations of the prelocking algorithm,
in particular maintain lex->query_tables_own_last pointer
in sync with lex->query_tables_last pointer/the contents
of lex->query_tables. When adding merge children, the merge
engine would extend the table list. Then, when adding
prelocked tables, the prelocking algorithm would use a pointer
to the last merge child to assign to lex->query_tables_own_last.
Then, when merge children were removed at the end of
open_tables(), lex->query_tables_own_last
was not updated, and kept pointing
to a removed merge child.
The fix ensures that query_tables_own_last is always in
sync with lex->query_tables_last.
This is a regression introduced by WL#4144 and present only
in next-4284 tree and 6.0.
mysql-test/r/merge.result:
Update results (Bug#47648).
mysql-test/t/merge.test:
Add a test case for Bug#47648.
Update the result file to reflect a fix of another bug
in MyISAMMRG code: not maintaining lex->query_tables_own_last
allowed a stored function or trigger to modify a merge table
which was already updated by the main statement.
It is not allowed for other storage engines, and should
not be allowed for MyISAMMRG.
storage/myisammrg/ha_myisammrg.cc:
When adding children to the list of tables to open,
make sure that we properly set lex->query_tables_own_last.
When removing the children, update lex->query_tables_own_last
if necessary.
2010-02-15 17:35:53 +01:00
#
# A test case for Bug#47648 main.merge fails sporadically
#
# Make sure we correctly maintain lex->query_tables_last_own.
#
create table t1 (c1 int not null);
create table t2 (c1 int not null);
create table t3 (c1 int not null);
create function f1 () returns int return (select max(c1) from t3);
create table t4 (c1 int not null) engine=merge union=(t1,t2) insert_method=last ;
select * from t4 where c1 < f1();
c1
prepare stmt from "select * from t4 where c1 < f1()";
execute stmt;
c1
execute stmt;
c1
execute stmt;
c1
drop function f1;
execute stmt;
ERROR 42000: FUNCTION test.f1 does not exist
execute stmt;
ERROR 42000: FUNCTION test.f1 does not exist
drop table t4, t3, t2, t1;
2010-02-26 13:40:25 +01:00
#
# Bug#51240 ALTER TABLE of a locked MERGE table fails
#
DROP TABLE IF EXISTS m1, t1;
CREATE TABLE t1 (c1 INT);
CREATE TABLE m1 (c1 INT) ENGINE=MRG_MyISAM UNION=(t1);
LOCK TABLE m1 WRITE;
ALTER TABLE m1 ADD INDEX (c1);
UNLOCK TABLES;
DROP TABLE m1, t1;
2010-06-11 10:14:38 +02:00
#
2010-10-06 09:56:29 +02:00
# Locking the merge table won't implicitly lock children.
2010-06-25 17:32:59 +02:00
#
CREATE TABLE t1 (c1 INT);
CREATE TABLE m1 (c1 INT) ENGINE=MRG_MyISAM UNION=(t1);
LOCK TABLE m1 WRITE;
ALTER TABLE t1 ADD INDEX (c1);
ERROR HY000: Table 't1' was locked with a READ lock and can't be updated
LOCK TABLE m1 WRITE, t1 WRITE;
ALTER TABLE t1 ADD INDEX (c1);
UNLOCK TABLES;
DROP TABLE m1, t1;
#
2010-06-11 10:14:38 +02:00
# Test for bug #37371 "CREATE TABLE LIKE merge loses UNION parameter"
#
drop tables if exists t1, m1, m2;
create table t1 (i int) engine=myisam;
create table m1 (i int) engine=mrg_myisam union=(t1) insert_method=first;
create table m2 like m1;
# Table definitions should match
show create table m1;
Table Create Table
m1 CREATE TABLE `m1` (
`i` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=FIRST UNION=(`t1`)
show create table m2;
Table Create Table
m2 CREATE TABLE `m2` (
`i` int(11) DEFAULT NULL
) ENGINE=MRG_MyISAM DEFAULT CHARSET=latin1 INSERT_METHOD=FIRST UNION=(`t1`)
drop tables m1, m2, t1;
2010-06-29 16:01:33 +02:00
#
# Test case for Bug#54811 "Assert in mysql_lock_have_duplicate()"
# Check that unique_table() works correctly for merge tables.
#
drop table if exists t1, t2, t3, m1, m2;
create table t1 (a int);
create table t2 (a int);
create table t3 (b int);
create view v1 as select * from t3,t1;
create table m1 (a int) engine=merge union (t1, t2) insert_method=last;
create table m2 (a int) engine=merge union (t1, t2) insert_method=first;
create temporary table tmp (b int);
insert into tmp (b) values (1);
insert into t1 (a) values (1);
insert into t3 (b) values (1);
insert into m1 (a) values ((select max(a) from m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t3, m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t3, m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t3, t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from t3, t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from tmp, m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from tmp, m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from tmp, t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from tmp, t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
insert into m1 (a) values ((select max(a) from v1));
ERROR HY000: The definition of table 'v1' prevents operation INSERT on table 'm1'.
insert into m1 (a) values ((select max(a) from tmp, v1));
ERROR HY000: The definition of table 'v1' prevents operation INSERT on table 'm1'.
update m1 set a = ((select max(a) from m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t3, m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t3, m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t3, t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from t3, t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from tmp, m1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from tmp, m2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from tmp, t1));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from tmp, t2));
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
update m1 set a = ((select max(a) from v1));
ERROR HY000: The definition of table 'v1' prevents operation UPDATE on table 'm1'.
update m1 set a = ((select max(a) from tmp, v1));
ERROR HY000: The definition of table 'v1' prevents operation UPDATE on table 'm1'.
delete from m1 where a = (select max(a) from m1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from m2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t3, m1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t3, m2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t3, t1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from t3, t2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from tmp, m1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from tmp, m2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from tmp, t1);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from tmp, t2);
ERROR HY000: You can't specify target table 'm1' for update in FROM clause
delete from m1 where a = (select max(a) from v1);
ERROR HY000: The definition of table 'v1' prevents operation DELETE on table 'm1'.
delete from m1 where a = (select max(a) from tmp, v1);
ERROR HY000: The definition of table 'v1' prevents operation DELETE on table 'm1'.
drop view v1;
drop temporary table tmp;
drop table t1, t2, t3, m1, m2;
Bug #56494 Segfault in upgrade_shared_lock_to_exclusive() for
REPAIR of merge table
Bug #56422 CHECK TABLE run when the table is locked reports
corruption along with timeout
The crash happened if a table maintenance statement (ANALYZE TABLE,
REPAIR TABLE, etc.) was executed on a MERGE table and opening and
locking a child table failed. This could for example happen if a child
table did not exist or if a lock timeout happened while waiting for
a conflicting metadata lock to disappear.
Since opening and locking the MERGE table and its children failed,
the tables would be closed and the metadata locks released.
However, TABLE_LIST::table for the MERGE table would still be set,
with its value invalid since the tables had been closed.
This caused the table maintenance statement to try to continue
and upgrade the metadata lock on the MERGE table. But since the lock
already had been released, this caused a segfault.
This patch fixes the problem by setting TABLE_LIST::table to NULL
if open_and_lock_tables() fails. This prevents maintenance
statements from continuing and trying to upgrade the metadata lock.
The patch includes a 5.5 version of the fix for
Bug #46339 crash on REPAIR TABLE merge table USE_FRM.
This bug caused REPAIR TABLE ... USE_FRM to give an assert
when used on merge tables.
The patch also enables the CHECK TABLE statement for log tables.
Before, CHECK TABLE for log tables gave ER_CANT_LOCK_LOG_TABLE,
yet still counted the statement as successfully executed.
With the changes to table maintenance statement error handling
in this patch, CHECK TABLE would no longer be considered as
successful in this case. This would have caused upgrade scripts
to mistakenly think that the general and slow logs are corrupted
and have to be repaired. Enabling CHECK TABLES for log tables
prevents this from happening.
Finally, the patch changes the error message from "Corrupt" to
"Operation failed" for a number of issues not related to table
corruption. For example "Lock wait timeout exceeded" and
"Deadlock found trying to get lock".
Test cases added to merge.test and check.test.
2010-09-22 10:15:41 +02:00
#
# Bug#56494 Segfault in upgrade_shared_lock_to_exclusive() for
# REPAIR of merge table
#
DROP TABLE IF EXISTS t1, t2, t_not_exists;
CREATE TABLE t1(a INT);
ALTER TABLE t1 engine= MERGE UNION (t_not_exists);
ANALYZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 analyze Error Table 'test.t_not_exists' doesn't exist
test.t1 analyze Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.t1 analyze error Corrupt
CHECK TABLE t1;
Table Op Msg_type Msg_text
2011-07-22 14:31:10 +02:00
test.t1 check Error Table 'test.t_not_exists' is differently defined or of non-MyISAM type or doesn't exist
Bug #56494 Segfault in upgrade_shared_lock_to_exclusive() for
REPAIR of merge table
Bug #56422 CHECK TABLE run when the table is locked reports
corruption along with timeout
The crash happened if a table maintenance statement (ANALYZE TABLE,
REPAIR TABLE, etc.) was executed on a MERGE table and opening and
locking a child table failed. This could for example happen if a child
table did not exist or if a lock timeout happened while waiting for
a conflicting metadata lock to disappear.
Since opening and locking the MERGE table and its children failed,
the tables would be closed and the metadata locks released.
However, TABLE_LIST::table for the MERGE table would still be set,
with its value invalid since the tables had been closed.
This caused the table maintenance statement to try to continue
and upgrade the metadata lock on the MERGE table. But since the lock
already had been released, this caused a segfault.
This patch fixes the problem by setting TABLE_LIST::table to NULL
if open_and_lock_tables() fails. This prevents maintenance
statements from continuing and trying to upgrade the metadata lock.
The patch includes a 5.5 version of the fix for
Bug #46339 crash on REPAIR TABLE merge table USE_FRM.
This bug caused REPAIR TABLE ... USE_FRM to give an assert
when used on merge tables.
The patch also enables the CHECK TABLE statement for log tables.
Before, CHECK TABLE for log tables gave ER_CANT_LOCK_LOG_TABLE,
yet still counted the statement as successfully executed.
With the changes to table maintenance statement error handling
in this patch, CHECK TABLE would no longer be considered as
successful in this case. This would have caused upgrade scripts
to mistakenly think that the general and slow logs are corrupted
and have to be repaired. Enabling CHECK TABLES for log tables
prevents this from happening.
Finally, the patch changes the error message from "Corrupt" to
"Operation failed" for a number of issues not related to table
corruption. For example "Lock wait timeout exceeded" and
"Deadlock found trying to get lock".
Test cases added to merge.test and check.test.
2010-09-22 10:15:41 +02:00
test.t1 check Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.t1 check error Corrupt
CHECKSUM TABLE t1;
Table Checksum
test.t1 NULL
Warnings:
Error 1146 Table 'test.t_not_exists' doesn't exist
Error 1168 Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
OPTIMIZE TABLE t1;
Table Op Msg_type Msg_text
test.t1 optimize Error Table 'test.t_not_exists' doesn't exist
test.t1 optimize Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.t1 optimize error Corrupt
REPAIR TABLE t1;
Table Op Msg_type Msg_text
2011-07-22 14:31:10 +02:00
test.t1 repair Error Table 'test.t_not_exists' is differently defined or of non-MyISAM type or doesn't exist
Bug #56494 Segfault in upgrade_shared_lock_to_exclusive() for
REPAIR of merge table
Bug #56422 CHECK TABLE run when the table is locked reports
corruption along with timeout
The crash happened if a table maintenance statement (ANALYZE TABLE,
REPAIR TABLE, etc.) was executed on a MERGE table and opening and
locking a child table failed. This could for example happen if a child
table did not exist or if a lock timeout happened while waiting for
a conflicting metadata lock to disappear.
Since opening and locking the MERGE table and its children failed,
the tables would be closed and the metadata locks released.
However, TABLE_LIST::table for the MERGE table would still be set,
with its value invalid since the tables had been closed.
This caused the table maintenance statement to try to continue
and upgrade the metadata lock on the MERGE table. But since the lock
already had been released, this caused a segfault.
This patch fixes the problem by setting TABLE_LIST::table to NULL
if open_and_lock_tables() fails. This prevents maintenance
statements from continuing and trying to upgrade the metadata lock.
The patch includes a 5.5 version of the fix for
Bug #46339 crash on REPAIR TABLE merge table USE_FRM.
This bug caused REPAIR TABLE ... USE_FRM to give an assert
when used on merge tables.
The patch also enables the CHECK TABLE statement for log tables.
Before, CHECK TABLE for log tables gave ER_CANT_LOCK_LOG_TABLE,
yet still counted the statement as successfully executed.
With the changes to table maintenance statement error handling
in this patch, CHECK TABLE would no longer be considered as
successful in this case. This would have caused upgrade scripts
to mistakenly think that the general and slow logs are corrupted
and have to be repaired. Enabling CHECK TABLES for log tables
prevents this from happening.
Finally, the patch changes the error message from "Corrupt" to
"Operation failed" for a number of issues not related to table
corruption. For example "Lock wait timeout exceeded" and
"Deadlock found trying to get lock".
Test cases added to merge.test and check.test.
2010-09-22 10:15:41 +02:00
test.t1 repair Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.t1 repair error Corrupt
REPAIR TABLE t1 USE_FRM;
Table Op Msg_type Msg_text
test.t1 repair Warning Can't open table
test.t1 repair error Corrupt
DROP TABLE t1;
CREATE TABLE t1(a INT);
CREATE TABLE t2(a INT) engine= MERGE UNION (t1);
REPAIR TABLE t2 USE_FRM;
Table Op Msg_type Msg_text
test.t2 repair note The storage engine for the table doesn't support repair
DROP TABLE t1, t2;
2010-10-06 09:56:29 +02:00
#
# Bug#57002 Assert in upgrade_shared_lock_to_exclusive()
# for ALTER TABLE + MERGE tables
#
DROP TABLE IF EXISTS t1, m1;
CREATE TABLE t1(a INT) engine=myisam;
CREATE TABLE m1(a INT) engine=merge UNION(t1);
LOCK TABLES t1 READ, m1 WRITE;
ALTER TABLE t1 engine=myisam;
ERROR HY000: Table 't1' was locked with a READ lock and can't be updated
UNLOCK TABLES;
DROP TABLE m1, t1;
2011-07-22 14:31:10 +02:00
#
# Test for bug #11754210 - "45777: CHECK TABLE DOESN'T SHOW ALL
# PROBLEMS FOR MERGE TABLE COMPLIANCE IN 5.1"
#
drop tables if exists t1, t2, t3, t4, m1;
create table t1(id int) engine=myisam;
create view t3 as select 1 as id;
create table t4(id int) engine=memory;
create table m1(id int) engine=merge union=(t1,t2,t3,t4);
select * from m1;
ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
# The below CHECK and REPAIR TABLE statements should
# report all problems with underlying tables:
# - absence of 't2',
# - missing base table for 't3',
# - wrong engine of 't4'.
check table m1;
Table Op Msg_type Msg_text
test.m1 check Error Table 'test.t2' is differently defined or of non-MyISAM type or doesn't exist
test.m1 check Error Table 'test.t3' is differently defined or of non-MyISAM type or doesn't exist
test.m1 check Error Table 'test.t4' is differently defined or of non-MyISAM type or doesn't exist
test.m1 check Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.m1 check error Corrupt
repair table m1;
Table Op Msg_type Msg_text
test.m1 repair Error Table 'test.t2' is differently defined or of non-MyISAM type or doesn't exist
test.m1 repair Error Table 'test.t3' is differently defined or of non-MyISAM type or doesn't exist
test.m1 repair Error Table 'test.t4' is differently defined or of non-MyISAM type or doesn't exist
test.m1 repair Error Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
test.m1 repair error Corrupt
# Clean-up.
drop tables m1, t1, t4;
drop view t3;
End of 5.5 tests