The bug could cause choosing a sub-optimal execution plan for
a single-table query if a unique index with many null keys were
defined for the table.
It happened because the code of the check_quick_keys function
made an assumption that any key may occur in an unique index
only once. Yet this is not true for keys with nulls that may
have multiple occurrences in the index.
Two problems here:
Problem 1:
While constructing the join columns list the optimizer does as follows:
1. Sets the join_using_fields/natural_join members of the right JOIN
operand.
2. Makes a "table reference" (TABLE_LIST) to parent the two tables.
3. Assigns the join_using_fields/is_natural_join of the wrapper table
using join_using_fields/natural_join of the rightmost table
4. Sets join_using_fields to NULL for the right JOIN operand.
5. Passes the parent table up to the same procedure on the upper
level.
Step 1 overrides the the join_using_fields that are set for a nested
join wrapping table in step 4.
Fixed by making a designated variable SELECT_LEX::prev_join_using to
pass the data from step 1 to step 4 without destroying the wrapping
table data.
Problem 2:
The optimizer checks for ambiguous columns while transforming
NATURAL JOIN/JOIN USING to JOIN ON. While doing that there was no
distinction between columns that are used in the generated join
condition (where ambiguity can be checked) and the other columns
(where ambiguity can be checked only when resolving references
coming from outside the JOIN construct itself).
Fixed by allowing the non-USING columns to be present in multiple
copies in both sides of the join and moving the ambiguity check
to the place where unqualified references to the join columns are
resolved (find_field_in_natural_join()).
When a merge table is opened compare column and key definition of
underlying tables against column and key definition of merge table.
If any of underlying tables have different column/key definition
refuse to open merge table.
The optimizer takes away columns from GROUP BY/DISTINCT if they constitute
all the parts of an unique index.
However if some of the columns can contain NULLs this cannot be done
(because an UNIQUE index can have multiple rows with NULL values).
Fixed by not using UNIQUE indexes with nullable columns to remove
grouping columns from GROUP BY/DISTINCT.
Depending on the queries we use different data processing methods
and can lose some data in case of double (and decimal in 4.1) fields.
The fix consists of two parts:
1. double comparison changed, now double a is equal to double b
if (a-b) is less than 5*0.1^(1 + max(a->decimals, b->decimals)).
For example, if a->decimals==1, b->decimals==2, a==b if (a-b)<0.005
2. if we use a temporary table, store double values there as is
to avoid any data conversion (rounding).
If inserting a GPL header, include a complete one
Makefile.am, mysql.dsw, mysql.sln:
Removed C version of mysql-test-run
mysql.spec.sh:
Specify GPL license only in GPL sources
.del-my_manage.h:
Delete: mysql-test/my_manage.h
mysql.spec.sh:
Added GPL header
.del-mysql_test_run_new.c:
Delete: mysql-test/mysql_test_run_new.c
.del-mysql_test_run_new.dsp:
Delete: VC++Files/mysql-test/mysql_test_run_new.dsp
.del-my_create_tables.c:
Delete: mysql-test/my_create_tables.c
.del-mysql_test_run_new_ia64.dsp:
Delete: VC++Files/mysql-test/mysql_test_run_new_ia64.dsp
msql2mysql.sh:
Use up-to-date GPL header
.del-mysql_test_run_new.vcproj:
Delete: VC++Files/mysql-test/mysql_test_run_new.vcproj
.del-my_manage.c:
Delete: mysql-test/my_manage.c
Made the function opt_sum_query to return HA_ERR_KEY_NOT_FOUND when
no matches were found (instead of -1 it returned prior this patch).
This changes allow us to avoid possible conflicts with return values
from user-defined handler methods which also may return -1.
No particular test cases are provided with this fix.
Checking for NULL before calling the val_xxx()
methods only checks for such arguments that are
known to be NULLs at compile time.
The arguments that may or may not contain
NULLs (e.g. function calls and possibly others)
are not checked at all.
Fixed by first calling the val_xxx() method and
then checking for null in SEC_TO_TIME().
In addition QUARTER() was not returning 0 (as all the
val_int() functions do when processing a NULL value).
Objects of the classes Item_func_is_not_null_test and Item_func_trig_cond
must be transparent for the method Item::split_sum_func2 as these classes
are pure helpers. It means that the method Item::split_sum_func2 should
look at those objects as at pure wrappers.
variables
Bits higher than 2**31 were impossible to set on THD::options. It's
probably a remnant from a time when options was a 32-bit integer.
Now, use unsigned long-long constants and variables to set and clear
THD::options.
---
Pulled comments back to column 49.
The bug report has demonstrated the following two problems.
1. If an ORDER/GROUP BY list includes a constant expression being
optimized away and, at the same time, containing single-row
subselects that return more that one row, no error is reported.
Strictly speaking the standard allows to ignore error in this case.
Yet, now a corresponding fatal error is reported in this case.
2. If a query requires sorting by expressions containing single-row
subselects that, however, return more than one row, then the execution
of the query may cause a server crash.
To fix this some code has been added that blocks execution of a subselect
item in case of a fatal error in the method Item_subselect::exec.
