Below is the list of changes that have just been committed into a local
4.1 repository of kgeorge. When kgeorge does a push these changes will
be propagated to the main repository and, within 24 hours after the
push, to the public repository.
For information on how to access the public repository
see http://dev.mysql.com/doc/mysql/en/installing-source-tree.html
ChangeSet@stripped, 2006-07-26 19:19:30+03:00, gkodinov@stripped +3 -0
* Bug #20792: Incorrect results from aggregate subquery
When processing aggregate functions all tables values are reset
to NULLs at the end of each group.
When doing that if there are no rows found for a group
the const tables must not be reset as they are not recalculated
by do_select()/sub_select() for each group.
mysql-test/r/subselect2.result@stripped, 2006-07-26 19:19:23+03:00, gkodinov@stripped +12 -0
* Bug #20792: Incorrect results from aggregate subquery
- test suite for the bug. This is dependent on InnoDB despite
the fact that the bug and the fix are not InnoDB specific.
This is because of the table flag HA_NOT_EXACT_COUNT.
When this flag is off (as in MyISAM) both t2 and t3 become of
join type 'system' as they are estimated to have 1 record and
and this statistics can be trusted (according to the absence of
HA_NOT_EXACT_COUNT).
mysql-test/t/subselect2.test@stripped, 2006-07-26 19:19:23+03:00, gkodinov@stripped +18 -0
* Bug #20792: Incorrect results from aggregate subquery
- test suite for the bug
sql/sql_select.cc@stripped, 2006-07-26 19:19:24+03:00, gkodinov@stripped +8 -2
* Bug #20792: Incorrect results from aggregate subquery
- when clearing results if there are not rows found for group
the const tables must not be reset as they are not recalculated
for each group.
# This is a BitKeeper patch. What follows are the unified diffs for the
# set of deltas contained in the patch. The rest of the patch, the part
# that BitKeeper cares about, is below these diffs.
# User: gkodinov
# Host: macbook.gmz
# Root: /Users/kgeorge/mysql/work/B20792-4.1-opt
--- 1.460/sql/sql_select.cc 2006-07-26 19:19:42 +03:00
+++ 1.461/sql/sql_select.cc 2006-07-26 19:19:42 +03:00
@@ -4452,10 +4452,16 @@ return_zero_rows(JOIN *join, select_resu
DBUG_RETURN(0);
}
-
+/*
+ used only in JOIN::clear
+*/
static void clear_tables(JOIN *join)
{
- for (uint i=0 ; i < join->tables ; i++)
+ /*
+ must clear only the non-const tables, as const tables
+ are not re-calculated.
+ */
+ for (uint i=join->const_tables ; i < join->tables ; i++)
mark_as_null_row(join->table[i]); // All fields are NULL
}
--- 1.8/mysql-test/r/subselect2.result 2006-07-26 19:19:42 +03:00
+++ 1.9/mysql-test/r/subselect2.result 2006-07-26 19:19:42 +03:00
@@ -130,3 +130,15 @@ id select_type table type possible_keys
5 DEPENDENT SUBQUERY t3 unique_subquery PRIMARY,FFOLDERID_IDX PRIMARY 32 func 1 Using index; Using where
6 DEPENDENT SUBQUERY t3 unique_subquery PRIMARY,FFOLDERID_IDX,CMFLDRPARNT_IDX PRIMARY 32 func 1 Using index; Using where
drop table t1, t2, t3, t4;
+CREATE TABLE t1 (a int(10) , PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t1 VALUES (1),(2);
+CREATE TABLE t2 (a int(10), PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t2 VALUES (1);
+CREATE TABLE t3 (a int(10), b int(10), c int(10),
+PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t3 VALUES (1,2,1);
+SELECT t1.* FROM t1 WHERE (SELECT COUNT(*) FROM t3,t2 WHERE t3.c=t2.a
+and t2.a='1' AND t1.a=t3.b) > 0;
+a
+2
+DROP TABLE t1,t2,t3;
--- 1.6/mysql-test/t/subselect2.test 2006-07-26 19:19:42 +03:00
+++ 1.7/mysql-test/t/subselect2.test 2006-07-26 19:19:42 +03:00
@@ -150,3 +150,21 @@ EXPLAIN SELECT t2.*, t4.DOCTYPENAME, t1.
drop table t1, t2, t3, t4;
# End of 4.1 tests
+
+#
+# Bug #20792: Incorrect results from aggregate subquery
+#
+CREATE TABLE t1 (a int(10) , PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t1 VALUES (1),(2);
+
+CREATE TABLE t2 (a int(10), PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t2 VALUES (1);
+
+CREATE TABLE t3 (a int(10), b int(10), c int(10),
+ PRIMARY KEY (a)) Engine=InnoDB;
+INSERT INTO t3 VALUES (1,2,1);
+
+SELECT t1.* FROM t1 WHERE (SELECT COUNT(*) FROM t3,t2 WHERE t3.c=t2.a
+ and t2.a='1' AND t1.a=t3.b) > 0;
+
+DROP TABLE t1,t2,t3;
| Thread |
|---|
| • bk commit into 4.1 tree (gkodinov:1.2535) BUG#20792 | kgeorge | 26 Jul |