2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
# Check that binlog is ok when a transaction mixes updates to InnoDB and
|
2003-09-09 20:06:50 +03:00
|
|
|
# MyISAM.
|
|
|
|
# It would be nice to make this a replication test, but in 4.0 the
|
|
|
|
# slave is always with --skip-innodb in the testsuite. I (Guilhem) however
|
|
|
|
# did some tests manually on a slave; tables are replicated fine and
|
2003-11-20 13:07:25 -06:00
|
|
|
# Exec_Master_Log_Pos advances as expected.
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2007-06-07 19:18:01 +02:00
|
|
|
-- source include/have_log_bin.inc
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
-- source include/have_innodb.inc
|
2008-02-27 15:18:27 +01:00
|
|
|
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
--disable_warnings
|
|
|
|
drop table if exists t1, t2;
|
|
|
|
--enable_warnings
|
|
|
|
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
connect (con1,localhost,root,,);
|
|
|
|
connect (con2,localhost,root,,);
|
|
|
|
|
|
|
|
connection con1;
|
2003-12-10 04:31:42 +00:00
|
|
|
create table t1 (a int) engine=innodb;
|
|
|
|
create table t2 (a int) engine=myisam;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(1);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
commit;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(2);
|
|
|
|
insert into t2 select * from t1;
|
|
|
|
# should say some changes to non-transact1onal tables couldn't be rolled back
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(3);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(4);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback to savepoint my_savepoint;
|
|
|
|
commit;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(5);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(6);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback to savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(7);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
commit;
|
2003-09-09 20:06:50 +03:00
|
|
|
select a from t1 order by a; # check that savepoints work :)
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
|
|
|
# and when ROLLBACK is not explicit?
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
select get_lock("a",10);
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(8);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
disconnect con1;
|
|
|
|
|
|
|
|
connection con2;
|
|
|
|
# We want to SHOW BINLOG EVENTS, to know what was logged. But there is no
|
|
|
|
# guarantee that logging of the terminated con1 has been done yet (it may not
|
|
|
|
# even be started, so con1 may have not even attempted to lock the binlog yet;
|
|
|
|
# so SHOW BINLOG EVENTS may come before con1 does the loggin. To be sure that
|
|
|
|
# logging has been done, we use a user lock.
|
|
|
|
select get_lock("a",10);
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
# and when not in a transact1on?
|
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(9);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
# Check that when the query updat1ng the MyISAM table is the first in the
|
2003-12-20 00:38:30 +01:00
|
|
|
# transaction, we log it immediately.
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(10); # first make t1 non-empty
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t2 select * from t1;
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(11);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
commit;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
# Check that things work like before this BEGIN/ROLLBACK code was added,
|
|
|
|
# when t2 is INNODB
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-12-10 04:31:42 +00:00
|
|
|
alter table t2 engine=INNODB;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(12);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
commit;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(13);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(14);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(15);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback to savepoint my_savepoint;
|
|
|
|
commit;
|
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
reset master;
|
|
|
|
|
|
|
|
begin;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(16);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(17);
|
|
|
|
insert into t2 select * from t1;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
rollback to savepoint my_savepoint;
|
2003-09-09 20:06:50 +03:00
|
|
|
insert into t1 values(18);
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
commit;
|
2003-09-09 20:06:50 +03:00
|
|
|
select a from t1 order by a; # check that savepoints work :)
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2 minor edits, plus
fix for BUG#1113 "INSERT into non-trans table SELECT ; ROLLBACK" does not send warning"
and
fix for BUG#873 "In transaction, INSERT to non-trans table is written too early to binlog".
Now we don't always write the non-trans update immediately to the binlog;
if there is something in the binlog cache we write it to the binlog cache
(because the non-trans update could depend on a trans table which was modified
earlier in the transaction); then in case of ROLLBACK, we write the binlog
cache to the binlog, wrapped with BEGIN/ROLLBACK.
This guarantees that the slave does the same updates.
For ROLLBACK TO SAVEPOINT: when we execute a SAVEPOINT command we write it
to the binlog cache. At ROLLBACK TO SAVEPOINT, if some non-trans table was updated,
we write ROLLBACK TO SAVEPOINT to the binlog cache; when the transaction
terminates (COMMIT/ROLLBACK), the binlog cache will be flushed to the binlog
(because of the non-trans update) so we'll have SAVEPOINT and ROLLBACK TO
SAVEPOINT in the binlog.
Apart from this rare case of updates of mixed table types in transaction, the
usual way is still clear the binlog cache at ROLLBACK, or chop it at
ROLLBACK TO SAVEPOINT (meaning the SAVEPOINT command is also chopped, which
is fine).
Note that BUG#873 encompasses subbugs 1) and 2) of BUG#333 "3 binlogging bugs when doing INSERT with mixed InnoDB/MyISAM".
2003-08-22 15:39:24 +02:00
|
|
|
|
2004-11-04 19:19:23 +01:00
|
|
|
# Test for BUG#5714, where a MyISAM update in the transaction used to
|
|
|
|
# release row-level locks in InnoDB
|
|
|
|
|
|
|
|
connect (con3,localhost,root,,);
|
|
|
|
|
|
|
|
connection con3;
|
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
|
|
|
--disable_warnings
|
2006-02-12 13:26:30 -08:00
|
|
|
alter table t2 engine=MyISAM;
|
2004-11-04 19:19:23 +01:00
|
|
|
--enable_warnings
|
|
|
|
insert into t1 values (1);
|
|
|
|
begin;
|
|
|
|
select * from t1 for update;
|
|
|
|
|
|
|
|
connection con2;
|
|
|
|
select (@before:=unix_timestamp())*0; # always give repeatable output
|
|
|
|
begin;
|
|
|
|
send select * from t1 for update;
|
|
|
|
|
|
|
|
connection con3;
|
|
|
|
insert into t2 values (20);
|
|
|
|
|
|
|
|
connection con2;
|
|
|
|
--error 1205
|
|
|
|
reap;
|
|
|
|
select (@after:=unix_timestamp())*0; # always give repeatable output
|
|
|
|
# verify that innodb_lock_wait_timeout was exceeded. When there was
|
|
|
|
# the bug, the reap would return immediately after the insert into t2.
|
|
|
|
select (@after-@before) >= 2;
|
|
|
|
|
2003-09-09 20:06:50 +03:00
|
|
|
drop table t1,t2;
|
2005-11-15 13:38:06 -07:00
|
|
|
commit;
|
|
|
|
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
# test for BUG#7947 - DO RELEASE_LOCK() not written to binlog on rollback in
|
|
|
|
# the middle of a transaction
|
2005-11-15 13:38:06 -07:00
|
|
|
|
|
|
|
connection con2;
|
|
|
|
begin;
|
|
|
|
create temporary table ti (a int) engine=innodb;
|
|
|
|
rollback;
|
|
|
|
insert into ti values(1);
|
|
|
|
set autocommit=0;
|
2005-11-16 21:17:38 -07:00
|
|
|
create temporary table t1 (a int) engine=myisam;
|
|
|
|
commit;
|
|
|
|
insert t1 values (1);
|
|
|
|
rollback;
|
2005-11-15 13:38:06 -07:00
|
|
|
create table t0 (n int);
|
|
|
|
insert t0 select * from t1;
|
|
|
|
set autocommit=1;
|
|
|
|
insert into t0 select GET_LOCK("lock1",null);
|
|
|
|
set autocommit=0;
|
|
|
|
create table t2 (n int) engine=innodb;
|
|
|
|
insert into t2 values (3);
|
|
|
|
disconnect con2;
|
|
|
|
connection con3;
|
2005-11-16 21:17:38 -07:00
|
|
|
select get_lock("lock1",60);
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
2005-11-15 13:38:06 -07:00
|
|
|
do release_lock("lock1");
|
|
|
|
drop table t0,t2;
|
|
|
|
|
2005-07-28 03:22:47 +03:00
|
|
|
# End of 4.1 tests
|
2006-02-18 17:19:16 +01:00
|
|
|
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
#
|
|
|
|
# Test behaviour of CREATE ... SELECT when mixing MyISAM and InnoDB tables
|
|
|
|
#
|
|
|
|
|
|
|
|
set autocommit=0;
|
|
|
|
CREATE TABLE t1 (a int, b int) engine=myisam;
|
|
|
|
reset master;
|
|
|
|
INSERT INTO t1 values (1,1),(1,2);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TABLE t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
# This should give warning
|
|
|
|
DROP TABLE if exists t2;
|
|
|
|
INSERT INTO t1 values (3,3);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TEMPORARY TABLE t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
ROLLBACK;
|
|
|
|
# This should give warning
|
|
|
|
DROP TABLE IF EXISTS t2;
|
|
|
|
|
|
|
|
CREATE TABLE t2 (a int, b int, primary key (a)) engine=innodb;
|
|
|
|
INSERT INTO t1 VALUES (4,4);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TABLE IF NOT EXISTS t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
SELECT * from t2;
|
|
|
|
TRUNCATE table t2;
|
|
|
|
INSERT INTO t1 VALUES (5,5);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
INSERT INTO t2 select * from t1;
|
|
|
|
SELECT * FROM t2;
|
|
|
|
DROP TABLE t2;
|
|
|
|
|
|
|
|
INSERT INTO t1 values (6,6);
|
|
|
|
CREATE TEMPORARY TABLE t2 (a int, b int, primary key (a)) engine=innodb ;
|
|
|
|
INSERT INTO t1 values (7,7);
|
|
|
|
ROLLBACK;
|
|
|
|
INSERT INTO t1 values (8,8);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TEMPORARY TABLE IF NOT EXISTS t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
COMMIT;
|
|
|
|
INSERT INTO t1 values (9,9);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TEMPORARY TABLE IF NOT EXISTS t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
ROLLBACK;
|
|
|
|
SELECT * from t2;
|
|
|
|
TRUNCATE table t2;
|
|
|
|
INSERT INTO t1 values (10,10);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
INSERT INTO t2 select * from t1;
|
|
|
|
SELECT * from t1;
|
|
|
|
INSERT INTO t2 values (100,100);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TEMPORARY TABLE IF NOT EXISTS t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
COMMIT;
|
|
|
|
INSERT INTO t2 values (101,101);
|
2007-06-06 10:57:07 -07:00
|
|
|
--error ER_DUP_ENTRY
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
CREATE TEMPORARY TABLE IF NOT EXISTS t2 (primary key (a)) engine=innodb select * from t1;
|
|
|
|
ROLLBACK;
|
|
|
|
SELECT * from t2;
|
|
|
|
DROP TABLE t1,t2;
|
2007-03-29 21:38:03 +02:00
|
|
|
source include/show_binlog_events.inc;
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
|
2006-02-18 17:19:16 +01:00
|
|
|
# Test for BUG#16559 (ROLLBACK should always have a zero error code in
|
|
|
|
# binlog). Has to be here and not earlier, as the SELECTs influence
|
|
|
|
# XIDs differently between normal and ps-protocol (and SHOW BINLOG
|
|
|
|
# EVENTS above read XIDs).
|
|
|
|
|
|
|
|
connect (con4,localhost,root,,);
|
|
|
|
connection con3;
|
|
|
|
reset master;
|
|
|
|
create table t1 (a int) engine=innodb;
|
|
|
|
create table t2 (a int) engine=myisam;
|
|
|
|
select get_lock("a",10);
|
|
|
|
begin;
|
|
|
|
insert into t1 values(8);
|
|
|
|
insert into t2 select * from t1;
|
|
|
|
disconnect con3;
|
|
|
|
|
|
|
|
connection con4;
|
|
|
|
select get_lock("a",10); # wait for rollback to finish
|
2008-01-11 12:35:35 +01:00
|
|
|
flush logs;
|
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes
Changes that requires code changes in other code of other storage engines.
(Note that all changes are very straightforward and one should find all issues
by compiling a --debug build and fixing all compiler errors and all
asserts in field.cc while running the test suite),
- New optional handler function introduced: reset()
This is called after every DML statement to make it easy for a handler to
statement specific cleanups.
(The only case it's not called is if force the file to be closed)
- handler::extra(HA_EXTRA_RESET) is removed. Code that was there before
should be moved to handler::reset()
- table->read_set contains a bitmap over all columns that are needed
in the query. read_row() and similar functions only needs to read these
columns
- table->write_set contains a bitmap over all columns that will be updated
in the query. write_row() and update_row() only needs to update these
columns.
The above bitmaps should now be up to date in all context
(including ALTER TABLE, filesort()).
The handler is informed of any changes to the bitmap after
fix_fields() by calling the virtual function
handler::column_bitmaps_signal(). If the handler does caching of
these bitmaps (instead of using table->read_set, table->write_set),
it should redo the caching in this code. as the signal() may be sent
several times, it's probably best to set a variable in the signal
and redo the caching on read_row() / write_row() if the variable was
set.
- Removed the read_set and write_set bitmap objects from the handler class
- Removed all column bit handling functions from the handler class.
(Now one instead uses the normal bitmap functions in my_bitmap.c instead
of handler dedicated bitmap functions)
- field->query_id is removed. One should instead instead check
table->read_set and table->write_set if a field is used in the query.
- handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and
handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now
instead use table->read_set to check for which columns to retrieve.
- If a handler needs to call Field->val() or Field->store() on columns
that are not used in the query, one should install a temporary
all-columns-used map while doing so. For this, we provide the following
functions:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set);
field->val();
dbug_tmp_restore_column_map(table->read_set, old_map);
and similar for the write map:
my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set);
field->val();
dbug_tmp_restore_column_map(table->write_set, old_map);
If this is not done, you will sooner or later hit a DBUG_ASSERT
in the field store() / val() functions.
(For not DBUG binaries, the dbug_tmp_restore_column_map() and
dbug_tmp_restore_column_map() are inline dummy functions and should
be optimized away be the compiler).
- If one needs to temporary set the column map for all binaries (and not
just to avoid the DBUG_ASSERT() in the Field::store() / Field::val()
methods) one should use the functions tmp_use_all_columns() and
tmp_restore_column_map() instead of the above dbug_ variants.
- All 'status' fields in the handler base class (like records,
data_file_length etc) are now stored in a 'stats' struct. This makes
it easier to know what status variables are provided by the base
handler. This requires some trivial variable names in the extra()
function.
- New virtual function handler::records(). This is called to optimize
COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true.
(stats.records is not supposed to be an exact value. It's only has to
be 'reasonable enough' for the optimizer to be able to choose a good
optimization path).
- Non virtual handler::init() function added for caching of virtual
constants from engine.
- Removed has_transactions() virtual method. Now one should instead return
HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support
transactions.
- The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument
that is to be used with 'new handler_name()' to allocate the handler
in the right area. The xxxx_create_handler() function is also
responsible for any initialization of the object before returning.
For example, one should change:
static handler *myisam_create_handler(TABLE_SHARE *table)
{
return new ha_myisam(table);
}
->
static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root)
{
return new (mem_root) ha_myisam(table);
}
- New optional virtual function: use_hidden_primary_key().
This is called in case of an update/delete when
(table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined
but we don't have a primary key. This allows the handler to take precisions
in remembering any hidden primary key to able to update/delete any
found row. The default handler marks all columns to be read.
- handler::table_flags() now returns a ulonglong (to allow for more flags).
- New/changed table_flags()
- HA_HAS_RECORDS Set if ::records() is supported
- HA_NO_TRANSACTIONS Set if engine doesn't support transactions
- HA_PRIMARY_KEY_REQUIRED_FOR_DELETE
Set if we should mark all primary key columns for
read when reading rows as part of a DELETE
statement. If there is no primary key,
all columns are marked for read.
- HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some
cases (based on table->read_set)
- HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS
Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION.
- HA_DUPP_POS Renamed to HA_DUPLICATE_POS
- HA_REQUIRES_KEY_COLUMNS_FOR_DELETE
Set this if we should mark ALL key columns for
read when when reading rows as part of a DELETE
statement. In case of an update we will mark
all keys for read for which key part changed
value.
- HA_STATS_RECORDS_IS_EXACT
Set this if stats.records is exact.
(This saves us some extra records() calls
when optimizing COUNT(*))
- Removed table_flags()
- HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if
handler::records() gives an exact count() and
HA_STATS_RECORDS_IS_EXACT if stats.records is exact.
- HA_READ_RND_SAME Removed (no one supported this one)
- Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk()
- Renamed handler::dupp_pos to handler::dup_pos
- Removed not used variable handler::sortkey
Upper level handler changes:
- ha_reset() now does some overall checks and calls ::reset()
- ha_table_flags() added. This is a cached version of table_flags(). The
cache is updated on engine creation time and updated on open.
MySQL level changes (not obvious from the above):
- DBUG_ASSERT() added to check that column usage matches what is set
in the column usage bit maps. (This found a LOT of bugs in current
column marking code).
- In 5.1 before, all used columns was marked in read_set and only updated
columns was marked in write_set. Now we only mark columns for which we
need a value in read_set.
- Column bitmaps are created in open_binary_frm() and open_table_from_share().
(Before this was in table.cc)
- handler::table_flags() calls are replaced with handler::ha_table_flags()
- For calling field->val() you must have the corresponding bit set in
table->read_set. For calling field->store() you must have the
corresponding bit set in table->write_set. (There are asserts in
all store()/val() functions to catch wrong usage)
- thd->set_query_id is renamed to thd->mark_used_columns and instead
of setting this to an integer value, this has now the values:
MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE
Changed also all variables named 'set_query_id' to mark_used_columns.
- In filesort() we now inform the handler of exactly which columns are needed
doing the sort and choosing the rows.
- The TABLE_SHARE object has a 'all_set' column bitmap one can use
when one needs a column bitmap with all columns set.
(This is used for table->use_all_columns() and other places)
- The TABLE object has 3 column bitmaps:
- def_read_set Default bitmap for columns to be read
- def_write_set Default bitmap for columns to be written
- tmp_set Can be used as a temporary bitmap when needed.
The table object has also two pointer to bitmaps read_set and write_set
that the handler should use to find out which columns are used in which way.
- count() optimization now calls handler::records() instead of using
handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true).
- Added extra argument to Item::walk() to indicate if we should also
traverse sub queries.
- Added TABLE parameter to cp_buffer_from_ref()
- Don't close tables created with CREATE ... SELECT but keep them in
the table cache. (Faster usage of newly created tables).
New interfaces:
- table->clear_column_bitmaps() to initialize the bitmaps for tables
at start of new statements.
- table->column_bitmaps_set() to set up new column bitmaps and signal
the handler about this.
- table->column_bitmaps_set_no_signal() for some few cases where we need
to setup new column bitmaps but don't signal the handler (as the handler
has already been signaled about these before). Used for the momement
only in opt_range.cc when doing ROR scans.
- table->use_all_columns() to install a bitmap where all columns are marked
as use in the read and the write set.
- table->default_column_bitmaps() to install the normal read and write
column bitmaps, but not signaling the handler about this.
This is mainly used when creating TABLE instances.
- table->mark_columns_needed_for_delete(),
table->mark_columns_needed_for_delete() and
table->mark_columns_needed_for_insert() to allow us to put additional
columns in column usage maps if handler so requires.
(The handler indicates what it neads in handler->table_flags())
- table->prepare_for_position() to allow us to tell handler that it
needs to read primary key parts to be able to store them in
future table->position() calls.
(This replaces the table->file->ha_retrieve_all_pk function)
- table->mark_auto_increment_column() to tell handler are going to update
columns part of any auto_increment key.
- table->mark_columns_used_by_index() to mark all columns that is part of
an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow
it to quickly know that it only needs to read colums that are part
of the key. (The handler can also use the column map for detecting this,
but simpler/faster handler can just monitor the extra() call).
- table->mark_columns_used_by_index_no_reset() to in addition to other columns,
also mark all columns that is used by the given key.
- table->restore_column_maps_after_mark_index() to restore to default
column maps after a call to table->mark_columns_used_by_index().
- New item function register_field_in_read_map(), for marking used columns
in table->read_map. Used by filesort() to mark all used columns
- Maintain in TABLE->merge_keys set of all keys that are used in query.
(Simplices some optimization loops)
- Maintain Field->part_of_key_not_clustered which is like Field->part_of_key
but the field in the clustered key is not assumed to be part of all index.
(used in opt_range.cc for faster loops)
- dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map()
tmp_use_all_columns() and tmp_restore_column_map() functions to temporally
mark all columns as usable. The 'dbug_' version is primarily intended
inside a handler when it wants to just call Field:store() & Field::val()
functions, but don't need the column maps set for any other usage.
(ie:: bitmap_is_set() is never called)
- We can't use compare_records() to skip updates for handlers that returns
a partial column set and the read_set doesn't cover all columns in the
write set. The reason for this is that if we have a column marked only for
write we can't in the MySQL level know if the value changed or not.
The reason this worked before was that MySQL marked all to be written
columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden
bug'.
- open_table_from_share() does not anymore setup temporary MEM_ROOT
object as a thread specific variable for the handler. Instead we
send the to-be-used MEMROOT to get_new_handler().
(Simpler, faster code)
Bugs fixed:
- Column marking was not done correctly in a lot of cases.
(ALTER TABLE, when using triggers, auto_increment fields etc)
(Could potentially result in wrong values inserted in table handlers
relying on that the old column maps or field->set_query_id was correct)
Especially when it comes to triggers, there may be cases where the
old code would cause lost/wrong values for NDB and/or InnoDB tables.
- Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags:
OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG.
This allowed me to remove some wrong warnings about:
"Some non-transactional changed tables couldn't be rolled back"
- Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset
(thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose
some warnings about
"Some non-transactional changed tables couldn't be rolled back")
- Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table()
which could cause delete_table to report random failures.
- Fixed core dumps for some tests when running with --debug
- Added missing FN_LIBCHAR in mysql_rm_tmp_tables()
(This has probably caused us to not properly remove temporary files after
crash)
- slow_logs was not properly initialized, which could maybe cause
extra/lost entries in slow log.
- If we get an duplicate row on insert, change column map to read and
write all columns while retrying the operation. This is required by
the definition of REPLACE and also ensures that fields that are only
part of UPDATE are properly handled. This fixed a bug in NDB and
REPLACE where REPLACE wrongly copied some column values from the replaced
row.
- For table handler that doesn't support NULL in keys, we would give an error
when creating a primary key with NULL fields, even after the fields has been
automaticly converted to NOT NULL.
- Creating a primary key on a SPATIAL key, would fail if field was not
declared as NOT NULL.
Cleanups:
- Removed not used condition argument to setup_tables
- Removed not needed item function reset_query_id_processor().
- Field->add_index is removed. Now this is instead maintained in
(field->flags & FIELD_IN_ADD_INDEX)
- Field->fieldnr is removed (use field->field_index instead)
- New argument to filesort() to indicate that it should return a set of
row pointers (not used columns). This allowed me to remove some references
to sql_command in filesort and should also enable us to return column
results in some cases where we couldn't before.
- Changed column bitmap handling in opt_range.cc to be aligned with TABLE
bitmap, which allowed me to use bitmap functions instead of looping over
all fields to create some needed bitmaps. (Faster and smaller code)
- Broke up found too long lines
- Moved some variable declaration at start of function for better code
readability.
- Removed some not used arguments from functions.
(setup_fields(), mysql_prepare_insert_check_table())
- setup_fields() now takes an enum instead of an int for marking columns
usage.
- For internal temporary tables, use handler::write_row(),
handler::delete_row() and handler::update_row() instead of
handler::ha_xxxx() for faster execution.
- Changed some constants to enum's and define's.
- Using separate column read and write sets allows for easier checking
of timestamp field was set by statement.
- Remove calls to free_io_cache() as this is now done automaticly in ha_reset()
- Don't build table->normalized_path as this is now identical to table->path
(after bar's fixes to convert filenames)
- Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to
do comparision with the 'convert-dbug-for-diff' tool.
Things left to do in 5.1:
- We wrongly log failed CREATE TABLE ... SELECT in some cases when using
row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result)
Mats has promised to look into this.
- Test that my fix for CREATE TABLE ... SELECT is indeed correct.
(I added several test cases for this, but in this case it's better that
someone else also tests this throughly).
Lars has promosed to do this.
2006-06-04 18:52:22 +03:00
|
|
|
|
2008-03-03 10:17:32 +01:00
|
|
|
let $MYSQLD_DATADIR= `select @@datadir`;
|
2006-02-18 17:19:16 +01:00
|
|
|
# we check that the error code of the "ROLLBACK" event is 0 and not
|
|
|
|
# ER_SERVER_SHUTDOWN (i.e. disconnection just rolls back transaction
|
|
|
|
# and does not make slave to stop)
|
2008-01-11 12:35:35 +01:00
|
|
|
if (`select @@binlog_format = 'ROW'`)
|
|
|
|
{
|
2009-09-29 14:16:23 +03:00
|
|
|
--exec $MYSQL_BINLOG --start-position=525 $MYSQLD_DATADIR/master-bin.000001 > $MYSQLTEST_VARDIR/tmp/mix_innodb_myisam_binlog.output
|
2008-01-11 12:35:35 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
if (`select @@binlog_format = 'STATEMENT' || @@binlog_format = 'MIXED'`)
|
|
|
|
{
|
2009-09-29 14:16:23 +03:00
|
|
|
--exec $MYSQL_BINLOG --start-position=556 $MYSQLD_DATADIR/master-bin.000001 > $MYSQLTEST_VARDIR/tmp/mix_innodb_myisam_binlog.output
|
2008-01-11 12:35:35 +01:00
|
|
|
}
|
|
|
|
|
2006-02-20 09:34:02 +01:00
|
|
|
--replace_result $MYSQLTEST_VARDIR MYSQLTEST_VARDIR
|
2006-02-18 17:19:16 +01:00
|
|
|
eval select
|
2006-02-20 09:34:02 +01:00
|
|
|
(@a:=load_file("$MYSQLTEST_VARDIR/tmp/mix_innodb_myisam_binlog.output"))
|
2006-02-18 17:19:16 +01:00
|
|
|
is not null;
|
|
|
|
--replace_result $MYSQL_TEST_DIR MYSQL_TEST_DIR
|
|
|
|
eval select
|
2008-02-01 13:18:27 +01:00
|
|
|
@a like "%#%error_code=0%ROLLBACK\\n/*!*/;%ROLLBACK /* added by mysqlbinlog */;%" OR
|
|
|
|
@a like "%#%error_code=0%ROLLBACK\\r\\n/*!*/;%ROLLBACK /* added by mysqlbinlog */;%",
|
2006-02-18 17:19:16 +01:00
|
|
|
@a not like "%#%error_code=%error_code=%";
|
2006-02-18 21:08:41 +01:00
|
|
|
drop table t1, t2;
|
2007-07-30 18:27:36 +03:00
|
|
|
|
|
|
|
#
|
|
|
|
# Bug #27417 thd->no_trans_update.stmt lost value inside of SF-exec-stack
|
|
|
|
# bug #28960 non-trans temp table changes with insert .. select
|
|
|
|
# not binlogged after rollback
|
|
|
|
#
|
|
|
|
# testing appearence of insert into temp_table in binlog.
|
|
|
|
# There are two branches of execution that require different setup.
|
|
|
|
|
|
|
|
## send_eof() branch
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
|
|
|
|
create temporary table tt (a int unique);
|
|
|
|
create table ti (a int) engine=innodb;
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# action
|
|
|
|
|
|
|
|
begin;
|
|
|
|
insert into ti values (1);
|
|
|
|
insert into ti values (2) ;
|
|
|
|
insert into tt select * from ti;
|
|
|
|
rollback;
|
|
|
|
|
|
|
|
# check
|
|
|
|
|
|
|
|
select count(*) from tt /* 2 */;
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc;
|
2007-07-30 18:27:36 +03:00
|
|
|
select count(*) from ti /* zero */;
|
|
|
|
insert into ti select * from tt;
|
|
|
|
select * from ti /* that is what slave would miss - a bug */;
|
|
|
|
|
|
|
|
|
|
|
|
## send_error() branch
|
|
|
|
delete from ti;
|
|
|
|
delete from tt where a=1;
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# action
|
|
|
|
|
|
|
|
begin;
|
|
|
|
insert into ti values (1);
|
|
|
|
insert into ti values (2) /* to make the dup error in the following */;
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
insert into tt select * from ti /* one affected and error */;
|
|
|
|
rollback;
|
|
|
|
|
|
|
|
# check
|
|
|
|
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc;
|
2007-07-30 18:27:36 +03:00
|
|
|
select count(*) from ti /* zero */;
|
|
|
|
insert into ti select * from tt;
|
|
|
|
select * from tt /* that is what otherwise slave missed - the bug */;
|
|
|
|
|
2008-01-11 12:35:35 +01:00
|
|
|
drop table ti, tt;
|
2007-07-30 18:27:36 +03:00
|
|
|
|
|
|
|
|
|
|
|
#
|
|
|
|
# Bug #27417 thd->no_trans_update.stmt lost value inside of SF-exec-stack
|
|
|
|
#
|
|
|
|
# Testing asserts: if there is a side effect of modifying non-transactional
|
|
|
|
# table thd->no_trans_update.stmt must be TRUE;
|
|
|
|
# the assert is active with debug build
|
|
|
|
#
|
|
|
|
|
|
|
|
--disable_warnings
|
|
|
|
drop function if exists bug27417;
|
|
|
|
drop table if exists t1,t2;
|
|
|
|
--enable_warnings
|
|
|
|
# side effect table
|
|
|
|
CREATE TABLE t1 (a int NOT NULL auto_increment primary key) ENGINE=MyISAM;
|
|
|
|
# target tables
|
|
|
|
CREATE TABLE t2 (a int NOT NULL auto_increment, PRIMARY KEY (a));
|
|
|
|
|
|
|
|
delimiter |;
|
2007-10-13 15:49:42 +03:00
|
|
|
create function bug27417(n int)
|
2007-07-30 18:27:36 +03:00
|
|
|
RETURNS int(11)
|
|
|
|
begin
|
|
|
|
insert into t1 values (null);
|
|
|
|
return n;
|
|
|
|
end|
|
|
|
|
delimiter ;|
|
|
|
|
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
|
|
|
|
insert into t2 values (bug27417(1));
|
|
|
|
insert into t2 select bug27417(2);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
insert into t2 values (bug27417(2));
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* only (!) with fixes for #23333 will show there is the query */;
|
2007-07-30 18:27:36 +03:00
|
|
|
select count(*) from t1 /* must be 3 */;
|
|
|
|
|
|
|
|
reset master;
|
|
|
|
select count(*) from t2;
|
|
|
|
delete from t2 where a=bug27417(3);
|
|
|
|
select count(*) from t2 /* nothing got deleted */;
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the query must be in regardless of #23333 */;
|
2007-07-30 18:27:36 +03:00
|
|
|
select count(*) from t1 /* must be 5 */;
|
|
|
|
|
|
|
|
--enable_info
|
|
|
|
delete t2 from t2 where t2.a=bug27417(100) /* must not affect t2 */;
|
|
|
|
--disable_info
|
|
|
|
select count(*) from t1 /* must be 7 */;
|
|
|
|
|
2007-08-21 15:16:55 +03:00
|
|
|
# function bug27417 remains for the following testing of bug#23333
|
2007-07-30 18:27:36 +03:00
|
|
|
drop table t1,t2;
|
|
|
|
|
2007-08-21 15:16:55 +03:00
|
|
|
#
|
|
|
|
# Bug#23333 using the patch (and the test) for bug#27471
|
2007-10-13 15:49:42 +03:00
|
|
|
#
|
2007-08-21 15:16:55 +03:00
|
|
|
# throughout the bug tests
|
|
|
|
# t1 - non-trans side effects gatherer;
|
|
|
|
# t2 - transactional table;
|
|
|
|
#
|
2007-10-13 15:49:42 +03:00
|
|
|
|
2007-08-21 15:16:55 +03:00
|
|
|
CREATE TABLE t1 (a int NOT NULL auto_increment primary key) ENGINE=MyISAM;
|
|
|
|
CREATE TABLE t2 (a int, PRIMARY KEY (a)) ENGINE=InnoDB;
|
2007-10-13 15:49:42 +03:00
|
|
|
CREATE TABLE t3 (a int, PRIMARY KEY (a), b int unique) ENGINE=MyISAM;
|
|
|
|
CREATE TABLE t4 (a int, PRIMARY KEY (a), b int unique) ENGINE=Innodb;
|
|
|
|
CREATE TABLE t5 (a int, PRIMARY KEY (a)) ENGINE=InnoDB;
|
2007-08-21 15:16:55 +03:00
|
|
|
|
|
|
|
|
|
|
|
#
|
|
|
|
# INSERT
|
|
|
|
#
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
|
|
|
|
insert into t2 values (1);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
insert into t2 values (bug27417(1));
|
|
|
|
|
|
|
|
# check
|
|
|
|
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-08-21 15:16:55 +03:00
|
|
|
select count(*) from t1 /* must be 1 */;
|
|
|
|
|
|
|
|
#
|
|
|
|
# INSERT SELECT
|
|
|
|
#
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
|
|
|
insert into t2 values (2);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
insert into t2 select bug27417(1) union select bug27417(2);
|
|
|
|
|
|
|
|
# check
|
|
|
|
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-08-21 15:16:55 +03:00
|
|
|
select count(*) from t1 /* must be 2 */;
|
|
|
|
|
|
|
|
#
|
2007-10-13 15:49:42 +03:00
|
|
|
# UPDATE inc multi-update
|
2007-08-21 15:16:55 +03:00
|
|
|
#
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
insert into t3 values (1,1),(2,3),(3,4);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
update t3 set b=b+bug27417(1);
|
|
|
|
|
|
|
|
# check
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-08-21 15:16:55 +03:00
|
|
|
select count(*) from t1 /* must be 2 */;
|
|
|
|
|
2007-10-13 15:49:42 +03:00
|
|
|
## multi_update::send_eof() branch
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t3;
|
|
|
|
delete from t4;
|
|
|
|
insert into t3 values (1,1);
|
|
|
|
insert into t4 values (1,1),(2,2);
|
|
|
|
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
UPDATE t4,t3 SET t4.a=t3.a + bug27417(1) /* top level non-ta table */;
|
|
|
|
|
|
|
|
# check
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-10-13 15:49:42 +03:00
|
|
|
select count(*) from t1 /* must be 4 */;
|
|
|
|
|
|
|
|
## send_error() branch of multi_update
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
delete from t3;
|
|
|
|
delete from t4;
|
|
|
|
insert into t3 values (1,1),(2,2);
|
|
|
|
insert into t4 values (1,1),(2,2);
|
|
|
|
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
UPDATE t3,t4 SET t3.a=t4.a + bug27417(1);
|
|
|
|
|
|
|
|
# check
|
|
|
|
select count(*) from t1 /* must be 1 */;
|
|
|
|
|
|
|
|
# cleanup
|
|
|
|
drop table t4;
|
|
|
|
|
2007-08-21 15:16:55 +03:00
|
|
|
|
|
|
|
#
|
2007-10-13 15:49:42 +03:00
|
|
|
# DELETE incl multi-delete
|
2007-08-21 15:16:55 +03:00
|
|
|
#
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
|
|
|
delete from t3;
|
|
|
|
insert into t2 values (1);
|
|
|
|
insert into t3 values (1,1);
|
|
|
|
create trigger trg_del before delete on t2 for each row
|
|
|
|
insert into t3 values (bug27417(1), 2);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
delete from t2;
|
|
|
|
# check
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-08-21 15:16:55 +03:00
|
|
|
select count(*) from t1 /* must be 1 */;
|
|
|
|
|
2007-10-13 15:49:42 +03:00
|
|
|
# cleanup
|
|
|
|
drop trigger trg_del;
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
delete from t2;
|
|
|
|
delete from t5;
|
|
|
|
create trigger trg_del_t2 after delete on t2 for each row
|
|
|
|
insert into t1 values (1);
|
|
|
|
insert into t2 values (2),(3);
|
|
|
|
insert into t5 values (1),(2);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
|
|
|
delete t2.* from t2,t5 where t2.a=t5.a + 1;
|
|
|
|
|
|
|
|
# check
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-10-13 15:49:42 +03:00
|
|
|
select count(*) from t1 /* must be 1 */;
|
|
|
|
|
2007-08-21 15:16:55 +03:00
|
|
|
|
|
|
|
#
|
|
|
|
# LOAD DATA
|
|
|
|
#
|
|
|
|
|
|
|
|
# prepare
|
|
|
|
delete from t1;
|
|
|
|
create table t4 (a int default 0, b int primary key) engine=innodb;
|
|
|
|
insert into t4 values (0, 17);
|
|
|
|
reset master;
|
|
|
|
|
|
|
|
# execute
|
|
|
|
--error ER_DUP_ENTRY
|
2008-03-03 10:17:32 +01:00
|
|
|
load data infile '../../std_data/rpl_loaddata.dat' into table t4 (a, @b) set b= @b + bug27417(2);
|
2007-08-21 15:16:55 +03:00
|
|
|
# check
|
|
|
|
select * from t4;
|
|
|
|
select count(*) from t1 /* must be 2 */;
|
2008-01-22 09:55:23 +08:00
|
|
|
source include/show_binlog_events.inc; /* the output must denote there is the query */;
|
2007-08-21 15:16:55 +03:00
|
|
|
|
|
|
|
#
|
|
|
|
# bug#23333 cleanup
|
|
|
|
#
|
|
|
|
|
|
|
|
|
2007-10-13 15:49:42 +03:00
|
|
|
drop trigger trg_del_t2;
|
|
|
|
drop table t1,t2,t3,t4,t5;
|
2007-08-21 15:16:55 +03:00
|
|
|
drop function bug27417;
|
|
|
|
|
|
|
|
|
2007-07-30 18:27:36 +03:00
|
|
|
--echo end of tests
|
|
|
|
|