To avoid unnecessary work the mysql_alter_table function takes the
list of table fields and applies all changes to it (drops/moves/renames/etc).
Then this function compares the new list and the old one. If the changes
require only .frm to be modified then the actual data isn't copied. To detect
changes all columns attributes but names are compared. When a column has been
moved and has replaced another column with the same attributes except name
the mysql_alter_table function wrongly decides that two fields has been just
renamed. As a result the data from the moved column and from all columns
after it is not copied.
Now the mysql_alter_table function forces table data copying by setting
the need_copy_table flag when it finds a moved column. The flag is set at
the stage when the modified fields are created.
sql/sql_table.cc:
Bug#28427: Columns were renamed instead of moving by ALTER TABLE.
Now the mysql_alter_table function forces table data copying by setting
the need_copy_table flag when it finds a moved column. The flag is set at
the stage when the modified fields are created.
mysql-test/r/alter_table.result:
Added a test case for the bug#28427: Columns were renamed instead of moving by ALTER TABLE.
mysql-test/t/alter_table.test:
Added a test case for the bug#28427: Columns were renamed instead of moving by ALTER TABLE.
To avoid unnecessary work the mysql_alter_table function takes the
list of table fields and applies all changes to it (drops/moves/renames/etc).
Then this function compares the new list and the old one. If the changes
require only .frm to be modified then the actual data isn't copied. To detect
changes all columns attributes but names are compared. When a column has been
moved and has replaced another column with the same attributes except name
the mysql_alter_table function wrongly decides that two fields has been just
renamed. As a result the data from the moved column and from all columns
after it is not copied.
Now the mysql_alter_table function forces table data copying by setting
the need_copy_table flag when it finds a moved column. The flag is set at
the stage when the modified fields are created.
into moonbone.local:/mnt/gentoo64/work/test-5.1-opt-mysql
mysql-test/r/func_gconcat.result:
Auto merged
mysql-test/t/func_gconcat.test:
Auto merged
sql/item_sum.cc:
Auto merged
sql/sql_base.cc:
Auto merged
sql/sql_select.cc:
Auto merged
sql/table.h:
Auto merged
mysql-test/r/alter_table.result:
Manual merge
mysql-test/t/alter_table.test:
Manual merge
sql/sql_table.cc:
Manual merge
mode.
When a new DATE/DATETIME field without default value is being added by the
ALTER TABLE the '0000-00-00' value is used as the default one. But it wasn't
checked whether such value was allowed by the set sql mode. Due to this
'0000-00-00' values was allowed for DATE/DATETIME fields even in the
NO_ZERO_DATE mode.
Now the mysql_alter_table() function checks whether the '0000-00-00' value
is allowed for DATE/DATETIME fields by the set sql mode.
The new error_if_not_empty flag is used in the mysql_alter_table() function
to indicate that it should abort if the table being altered isn't empty.
The new new_datetime_field field is used in the mysql_alter_table() function
for error throwing purposes.
The new error_if_not_empty parameter is added to the copy_data_between_tables()
function to indicate the it should return error if the source table isn't empty.
mysql-test/t/alter_table.test:
Added a test case for the bug#27507: Wrong DATETIME value was allowed by
ALTER TABLE in the NO_ZERO_DATE mode.
mysql-test/r/alter_table.result:
Added a test case for the bug#27507: Wrong DATETIME value was allowed by
ALTER TABLE in the NO_ZERO_DATE mode.
sql/sql_table.cc:
Bug#27507: Wrong DATETIME value was allowed by ALTER TABLE in the NO_ZERO_DATE
mode.
Now the mysql_alter_table() function checks whether the '0000-00-00' value
is allowed for DATE/DATETIME fields by the set sql mode.
The new error_if_not_empty flag is used in the mysql_alter_table() function
to indicate that it should abort if the table being altered isn't empty.
The new new_datetime_field field is used in the mysql_alter_table() function
for error throwing purposes.
The new error_if_not_empty parameter is added to the copy_data_between_tables()
function to indicate the it should return error if the source table isn't empty.
mode.
When a new DATE/DATETIME field without default value is being added by the
ALTER TABLE the '0000-00-00' value is used as the default one. But it wasn't
checked whether such value was allowed by the set sql mode. Due to this
'0000-00-00' values was allowed for DATE/DATETIME fields even in the
NO_ZERO_DATE mode.
Now the mysql_alter_table() function checks whether the '0000-00-00' value
is allowed for DATE/DATETIME fields by the set sql mode.
The new error_if_not_empty flag is used in the mysql_alter_table() function
to indicate that it should abort if the table being altered isn't empty.
The new new_datetime_field field is used in the mysql_alter_table() function
for error throwing purposes.
The new error_if_not_empty parameter is added to the copy_data_between_tables()
function to indicate the it should return error if the source table isn't empty.
and invalidation in the most general case (non-temporary table and
not simple RENAME or ENABLE/DISABLE KEYS or partitioning command).
See comment for sql/sql_table.cc for more information.
These changes are prerequisite for 5.1 version of fix for bug #23667
"CREATE TABLE LIKE is not isolated from alteration by other connections"
mysql-test/include/mix1.inc:
Extended coverage for behavior of ALTER TABLE statement under LOCK TABLES,
which should be consistent across all platforms and for all engines.
mysql-test/r/alter_table-big.result:
Changed test for bug #25044 to use @@debug and injected sleeps
infrastructure. Extended test coverage for ALTER TABLE's behavior
under concurrency.
mysql-test/r/alter_table.result:
Extended coverage for behavior of ALTER TABLE statement under LOCK TABLES,
which should be consistent across all platforms and for all engines.
mysql-test/r/innodb_mysql.result:
Extended coverage for behavior of ALTER TABLE statement under LOCK TABLES,
which should be consistent across all platforms and for all engines.
mysql-test/t/alter_table-big.test:
Changed test for bug #25044 to use @@debug and injected sleeps
infrastructure. Extended test coverage for ALTER TABLE's behavior
under concurrency.
mysql-test/t/alter_table.test:
Extended coverage for behavior of ALTER TABLE statement under LOCK TABLES,
which should be consistent across all platforms and for all engines.
sql/mysql_priv.h:
Made functions reopen_table() and close_handle_and_leave_table_as_lock()
available outside of sql_base.cc file.
Changed close_data_tables() in such way that after closing handler
for the table it leaves TABLE object for it in table cache not as
placeholder for ordinary name-lock but as placeholder for an exclusive
name-lock. Renamed this routine to close_data_files_and_morph_locks().
sql/sql_base.cc:
Made functions reopen_table() and close_handle_and_leave_table_as_lock()
available outside of sql_base.cc file.
Changed close_data_tables() in such way that after closing handler
for the table it leaves TABLE object for it in table cache not as
placeholder for ordinary name-lock but as placeholder for an exclusive
name-lock. Renamed this routine to close_data_files_and_morph_locks().
Also adjusted it so it can work properly not only in LOCK TABLES mode.
sql/sql_table.cc:
Changed the way in which ALTER TABLE implementation handles table
locking and invalidation in the most general case (non-temporary table
and not simple RENAME or ENABLE/DISABLE KEYS or partitioning command)
Now after preparing new version of the table we:
1) Wait until all other threads close old version of table.
2) Close instances of table open by this thread and replace them
with exclusive name-locks.
3) Rename the old table to a temp name, rename the new one to the
old name.
4) If we are under LOCK TABLES and don't do ALTER TABLE ... RENAME
we reopen new version of table.
5) Write statement to the binary log.
6) If we are under LOCK TABLES and do ALTER TABLE ... RENAME we
remove name-locks from list of open tables and table cache.
7) If we are not not under LOCK TABLES we rely on close_thread_tables()
call to remove name-locks from table cache and list of open table.
Such approach:
a) Eliminates possibility for concurrent statement to sneak in and get
access to the new version of the table before ALTER TABLE gets logged
into binary log.
b) Ensures that ALTER TABLE behaves under LOCK TABLES in the same way
on all platforms and for all engines (in 5.0 this was not true)
c) Preserves nice invariant that if table is open in some connection
there is a guarantee that .FRM file for this table exists and is
properly named.
and invalidation in the most general case (non-temporary table and
not simple RENAME or ENABLE/DISABLE KEYS or partitioning command).
See comment for sql/sql_table.cc for more information.
These changes are prerequisite for 5.1 version of fix for bug #23667
"CREATE TABLE LIKE is not isolated from alteration by other connections"
into pilot.blaudden:/home/msvensson/mysql/mysql-5.1-maint
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
sql/sql_table.cc:
Auto merged
into pilot.blaudden:/home/msvensson/mysql/mysql-5.1-maint
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
sql/sql_table.cc:
Auto merged
into pilot.blaudden:/home/msvensson/mysql/bug25262/my51-bug25262
mysql-test/r/alter_table.result:
Manual merge
mysql-test/t/alter_table.test:
Manual merge
sql/sql_table.cc:
Manual merge, auto_increment_value variable has moved into "stats"
It was syntactically correct to define
spatial keys over parts of columns (e.g.
ALTER TABLE t1 ADD x GEOMETRY NOT NULL,
ADD SPATIAL KEY (x(32))).
This may lead to undefined results and/or
interpretation.
Fixed by not allowing partial column
specification in a SPATIAL index definition.
mysql-test/r/alter_table.result:
Bug #26794: 5.1 part
test case
mysql-test/r/gis-rtree.result:
Bug #26794: 5.1 part
updated the tests to the new syntax
mysql-test/t/alter_table.test:
Bug #26794: 5.1 part
test case
mysql-test/t/gis-rtree.test:
Bug #26794: 5.1 part
updated the tests to the new syntax
sql/sql_table.cc:
Bug #26794: 5.1 part
Disable defining SPATIAL KEYS with sub-key parts
It was syntactically correct to define
spatial keys over parts of columns (e.g.
ALTER TABLE t1 ADD x GEOMETRY NOT NULL,
ADD SPATIAL KEY (x(32))).
This may lead to undefined results and/or
interpretation.
Fixed by not allowing partial column
specification in a SPATIAL index definition.
Different set of conditions is used to verify
the validity of index definitions over a GEOMETRY
column in ALTER TABLE and CREATE TABLE.
The difference was on how sub-keys notion validity
is checked.
Fixed by extending the CREATE TABLE condition to
support the cases allowed in ALTER TABLE.
Made the SHOW CREATE TABLE not to display spatial
indexes using the sub-key notion.
mysql-test/r/alter_table.result:
Bug #26794: test case
mysql-test/r/gis-rtree.result:
Bug #26794: fixed SHOW CREATE TABLE output.
mysql-test/t/alter_table.test:
Bug #26794: test case
sql/field.cc:
Bug #26794: Allow sub-keys for GEOMETRY
sql/sql_show.cc:
Bug #26794: Don't show sub-key notion
in SHOW CREATE TABLE for SPATIAL indexes.
sql/sql_table.cc:
Bug #26794: Allow sub-keys for GEOMETRY
Different set of conditions is used to verify
the validity of index definitions over a GEOMETRY
column in ALTER TABLE and CREATE TABLE.
The difference was on how sub-keys notion validity
is checked.
Fixed by extending the CREATE TABLE condition to
support the cases allowed in ALTER TABLE.
Made the SHOW CREATE TABLE not to display spatial
indexes using the sub-key notion.
into magare.gmz:/home/kgeorge/mysql/work/B26794-5.1-opt
mysql-test/r/gis-rtree.result:
Auto merged
sql/field.cc:
Auto merged
sql/sql_show.cc:
Auto merged
sql/sql_table.cc:
Auto merged
mysql-test/r/alter_table.result:
tests merge
mysql-test/t/alter_table.test:
tests merge
Different set of conditions is used to verify
the validity of index definitions over a GEOMETRY
column in ALTER TABLE and CREATE TABLE.
The difference was on how sub-keys notion validity
is checked.
Fixed by extending the CREATE TABLE condition to
support the cases allowed in ALTER TABLE.
Made the SHOW CREATE TABLE not to display spatial
indexes using the sub-key notion.
mysql-test/r/alter_table.result:
Bug #26794: test case
mysql-test/r/gis-rtree.result:
Bug #26794: fixed SHOW CREATE TABLE output.
mysql-test/t/alter_table.test:
Bug #26794: test case
sql/field.cc:
Bug #26794: Allow sub-keys for GEOMETRY
sql/sql_show.cc:
Bug #26794: Don't show sub-key notion
in SHOW CREATE TABLE for SPATIAL indexes.
sql/sql_table.cc:
Bug #26794: Allow sub-keys for GEOMETRY
Different set of conditions is used to verify
the validity of index definitions over a GEOMETRY
column in ALTER TABLE and CREATE TABLE.
The difference was on how sub-keys notion validity
is checked.
Fixed by extending the CREATE TABLE condition to
support the cases allowed in ALTER TABLE.
Made the SHOW CREATE TABLE not to display spatial
indexes using the sub-key notion.
- Try to copy the autoincrement value when altering the table
mysql-test/r/alter_table.result:
Update test result
mysql-test/t/alter_table.test:
Add test case as described in bug report
sql/sql_table.cc:
Try to copy the autoincrement value when altering the table
into weblab.(none):/home/marcsql/TREE/mysql-5.1-24562-merge
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
sql/sql_yacc.yy:
Auto merged
WL#3681 (ALTER TABLE ORDER BY)
Before this fix, the ALTER TABLE statement implemented an ORDER BY option
with the following characteristics :
1) The order by clause accepts a list of criteria, with optional ASC or
DESC keywords
2) Each criteria can be a general expression, involving operators,
native functions, stored functions, user defined functions, subselects ...
With this fix :
1) has been left unchanged, since it's a de-facto existing feature,
that was already present in the code base and partially covered in the test
suite. Code coverage for ASC and DESC was missing and has been improved.
2) has been changed to limit the kind of criteria that are permissible:
now only a column name is valid.
mysql-test/r/alter_table.result:
Prevent ALTER TABLE ORDER BY clauses to use general expressions.
mysql-test/t/alter_table.test:
Prevent ALTER TABLE ORDER BY clauses to use general expressions.
sql/sql_yacc.yy:
Prevent ALTER TABLE ORDER BY clauses to use general expressions.
WL#3681 (ALTER TABLE ORDER BY)
Before this fix, the ALTER TABLE statement implemented an ORDER BY option
with the following characteristics :
1) The order by clause accepts a list of criteria, with optional ASC or
DESC keywords
2) Each criteria can be a general expression, involving operators,
native functions, stored functions, user defined functions, subselects ...
With this fix :
1) has been left unchanged, since it's a de-facto existing feature,
that was already present in the code base and partially covered in the test
suite. Code coverage for ASC and DESC was missing and has been improved.
2) has been changed to limit the kind of criteria that are permissible:
now only a column name is valid.
into mysql.com:/home/svoj/devel/mysql/BUG23404/mysql-5.1-engines
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
sql/sql_parse.cc:
Auto merged
into mysql.com:/home/svoj/devel/mysql/BUG23404/mysql-5.0-engines
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
sql/sql_parse.cc:
SCCS merged
table
ROW_FORMAT option is lost during CREATE/DROP INDEX.
This fix forces CREATE/DROP INDEX to retain ROW_FORMAT by instructing
mysql_alter_table() that ROW_FORMAT is not used during creating/dropping
indexes.
mysql-test/r/alter_table.result:
A test case for bug#23404.
mysql-test/t/alter_table.test:
A test case for bug#23404.
sql/sql_parse.cc:
CREATE/DROP INDEX must not change ROW_FORMAT. Setting create_info.row_type
to ROW_TYPE_NOT_USED informs mysql_alter_table that ROW_FORMAT was not
used during alteration, and thus must be retained.
table
ROW_FORMAT option is lost during CREATE/DROP INDEX.
This fix forces CREATE/DROP INDEX to retain ROW_FORMAT by instructing
mysql_alter_table() that ROW_FORMAT is not used during creating/dropping
indexes.
into example.com:/work/bug22369-v2/my51
mysql-test/r/alter_table.result:
Auto merged
mysql-test/r/grant.result:
Auto merged
mysql-test/t/grant.test:
Auto merged
sql/mysql_priv.h:
Auto merged
sql/sql_parse.cc:
Auto merged
sql/sql_table.cc:
Auto merged
mysql-test/t/alter_table.test:
manual merge
with other alterations causes lost tables
Using RENAME clause combined with other clauses of ALTER TABLE led to
data loss (the data was there but not accessible). This could happen if the
changes do not change the table much. Adding and droppping of fields and
indices was safe. Renaming a column with MODIFY or CHANGE was unsafe operation,
if the actual column didn't change (changing from int to int, which is a noop)
Depending on the storage engine (SE) the behavior is different:
1)MyISAM/MEMORY - the ALTER TABLE statement completes
without any error but next SELECT against the new table fails.
2)InnoDB (and every other transactional table) - The ALTER TABLE statement
fails. There are the the following files in the db dir -
`new_table_name.frm` and a temporary table's frm. If the SE is file
based, then the data and index files will be present but with the old
names. What happens is that for InnoDB the table is not renamed in the
internal DDIC.
Fixed by adding additional call to mysql_rename_table() method, which should
not include FRM file rename, because it has been already done during file
names juggling.
mysql-test/r/alter_table.result:
update result
mysql-test/r/grant.result:
update result
mysql-test/t/alter_table.test:
2006/11/29 11:46:23+01:00 andrey@example.com +44 -9
Error to bug number
Added test case for #22369: Alter table rename combined
with other alterations causes lost tables
mysql-test/t/grant.test:
add test for bug#22369 - alter table was missing check
for DROP_ACL when ALTER_RENAME clause is specified. Synchronise
with RENAME TABLE DDL.
sql/mysql_priv.h:
Add a new flag for mysql_rename_table()
sql/sql_parse.cc:
To be consistent with SQLCOM_RENAME_TABLE, SQLCOM_ALTER_TABLE has
to check for DROP_ACL if there is ALTER_RENAME flag set.
sql/sql_table.cc:
ALTER_RENAME, the data and index files weren't renamed in the engine
but only the FRM was new, when the tables old and new tables are compatible.
In the chain of FRM renames we add a call to mysql_rename_table() which should
instruct the engine to rename the table but not rename the FRM.
This bug was there only in 5.1 branch. 4.1 and 5.0 always do copy data on RENAME
if there are more clauses than just rename.
with other alterations causes lost tables
Using RENAME clause combined with other clauses of ALTER TABLE led to
data loss (the data was there but not accessible). This could happen if the
changes do not change the table much. Adding and droppping of fields and
indices was safe. Renaming a column with MODIFY or CHANGE was unsafe operation,
if the actual column didn't change (changing from int to int, which is a noop)
Depending on the storage engine (SE) the behavior is different:
1)MyISAM/MEMORY - the ALTER TABLE statement completes
without any error but next SELECT against the new table fails.
2)InnoDB (and every other transactional table) - The ALTER TABLE statement
fails. There are the the following files in the db dir -
`new_table_name.frm` and a temporary table's frm. If the SE is file
based, then the data and index files will be present but with the old
names. What happens is that for InnoDB the table is not renamed in the
internal DDIC.
Fixed by adding additional call to mysql_rename_table() method, which should
not include FRM file rename, because it has been already done during file
names juggling.
into example.com:/work/bug24395-v2/my51
mysql-test/r/alter_table.result:
Auto merged
mysql-test/t/alter_table.test:
Auto merged
storage/myisam/mi_open.c:
Auto merged
sql/sql_table.cc:
manual merge
