Lines below which were added in the patch for Bug#56814 cause this crash:
+ if (table->table)
+ table->table->maybe_null= FALSE;
Consider following test case:
--
CREATE TABLE t1(f1 INT NOT NULL);
INSERT INTO t1 VALUES (16777214),(0);
SELECT COUNT(*) FROM t1 LEFT JOIN t1 t2
ON 1 WHERE t2.f1 > 1 GROUP BY t2.f1;
DROP TABLE t1;
--
We set TABLE::maybe_null to FALSE for t2 table
and in create_tmp_field() we create appropriate tmp table field
using create_tmp_field_from_item() function instead of
create_tmp_field_from_field. As a result we have
LONGLONG field. As we have GROUP BY clause we calculate
group buffer length, see calc_group_buffer().
Item from group list which is used for calculation
refer to the field from real tables and have LONG type.
So group buffer length become insufficient for storing of
LONGLONG value. It leads to overwriting of wrong memory
area in do_field_int() function which is called from
end_update().
After some investigation I found out that
create_tmp_field_from_item() is used only for OLAP
grouping and can not be used for common grouping
as it could be an incompatibility between tmp
table fields and group buffer length.
We can not remove create_tmp_field_from_item() call from
create_tmp_field as OLAP needs it and we can not use this
function for common grouping. So we should remove setting
TABLE::maybe_null to FALSE from simplify_joins().
In this case we'll get wrong behaviour of
list_contains_unique_index() back. To fix it we
could use Field::real_maybe_null() check instead of
Field::maybe_null() and add addition check of
TABLE_LIST::outer_join.
In order to be able to check if the set of the grouping fields in a
GROUP BY has changed (and thus to start a new group) the optimizer
caches the current values of these fields in a set of Cached_item
derived objects.
The Cached_item_str, used for caching varchar and TEXT columns,
is limited in length by the max_sort_length variable.
A String buffer to store the value with an alloced length of either
the max length of the string or the value of max_sort_length
(whichever is smaller) in Cached_item_str's constructor.
Then, at compare time the value of the string to compare to was
truncated to the alloced length of the string buffer inside
Cached_item_str.
This is all fine and valid, but only if you're not assigning
values near or equal to the alloced length of this buffer.
Because when assigning values like this the alloced length is
rounded up and as a result the next set of data will not match the
group buffer, thus leading to wrong results because of the changed
alloced_length.
Fixed by preserving the original maximum length in the
Cached_item_str's constructor and using this instead of the
alloced_length to limit the string to compare to.
Test case added.
This bug is a design flaw of the fix for the bug#33546. It assumed that an
item can be used only in one comparison context, but actually it isn't the
case. Item_cache_datetime is used to store result for MIX/MAX aggregate
functions. Because Arg_comparator always compares datetime values as INTs when
possible the Item_cache_datetime most time caches only INT value. But
since all datetime values has STRING result type MIN/MAX functions are asked
for a STRING value when the result is being sent to a client. The
Item_cache_datetime was designed to avoid conversions and get INT/STRING
values from an underlying item, but at the moment the values is asked
underlying item doesn't hold it anymore thus wrong result is returned.
Beside that MIN/MAX aggregate functions was wrongly initializing cached result
and this led to a wrong result.
The Item::has_compatible_context helper function is added. It checks whether
this and given items has the same comparison context or can be compared as
DATETIME values by Arg_comparator. The equality propagation optimization is
adjusted to take into account that items which being compared as DATETIME
can have different comparison contexts.
The Item_cache_datetime now converts cached INT value to a correct STRING
DATETIME value by means of number_to_datetime & my_TIME_to_str functions.
The Arg_comparator::set_cmp_context_for_datetime helper function is added.
It sets comparison context of items being compared as DATETIMEs to INT if
items will be compared as longlong.
The Item_sum_hybrid::setup function now correctly initializes its result
value.
In order to avoid unnecessary conversions Item_sum_hybrid now states that it
can provide correct longlong value if the item being aggregated can do it
too.
NULL from outer join query
Problem: optimising MIN/MAX() queries without GROUP BY clause
by replacing the aggregate expression with a constant, we may set it
to NULL disregarding the fact that there may be outer joins involved.
Fix: don't replace MIN/MAX() with NULL if there're outer joins.
Note: the fix itself is just
- if (!count)
+ if (!count && !outer_tables)
set to NULL
The rest of the patch eliminates repeated code to improve speed
and for easy maintenance of the code.
Grouping by a subquery in a query with a distinct aggregate
function lead to a wrong result (wrong and unordered
grouping values).
There are two related problems:
1) The query like this:
SELECT (SELECT t1.a) aa, COUNT(DISTINCT b) c
FROM t1 GROUP BY aa
returned wrong result, because the outer reference "t1.a"
in the subquery was substituted with the Item_ref item.
The Item_ref item obtains data from the result_field object
that refreshes once after the end of each group. This data
is not applicable to filesort since filesort() doesn't care
about groups (and doesn't update result_field objects with
copy_fields() and so on). Also that data is not applicable
to group separation algorithm: end_send_group() checks every
record with test_if_group_changed() that evaluates Item_ref
items, but it refreshes those Item_ref-s only after the end
of group, that is a vicious circle and the grouped column
values in the output are shifted.
Fix: if
a) we grouping by a subquery and
b) that subquery has outer references to FROM list
of the grouping query,
then we substitute these outer references with
Item_direct_ref like references under aggregate
functions: Item_direct_ref obtains data directly
from the current record.
2) The query with a non-trivial grouping expression like:
SELECT (SELECT t1.a) aa, COUNT(DISTINCT b) c
FROM t1 GROUP BY aa+0
also returned wrong result, since JOIN::exec() substitutes
references to top-level aliases in SELECT list with Item_copy
caching items. Item_copy items have same refreshing policy
as Item_ref items, so the whole groping expression with
Item_copy inside returns wrong result in filesort() and
end_send_group().
Fix: include aliased items into GROUP BY item tree instead
of Item_ref references to them.
------------------------------------------------------------
revno: 2476.784.3
committer: davi@moksha.local
timestamp: Tue 2007-10-02 21:27:31 -0300
message:
Bug#25858 Some DROP TABLE under LOCK TABLES can cause deadlocks
When a client (connection) holds a lock on a table and attempts to
drop (obtain a exclusive lock) on a second table that is already
held by a second client and the second client then attempts to
drop the table that is held by the first client, leads to a
circular wait deadlock. This scenario is very similar to trying to
drop (or rename) a table while holding read locks and are
correctly forbidden.
The solution is to allow a drop table operation to continue only
if the table being dropped is write (exclusively) locked, or if
the table is temporary, or if the client is not holding any
locks. Using this scheme prevents the creation of a circular
chain in which each client is waiting for one table that the
next client in the chain is holding.
This is incompatible change, as can be seen by number of tests
cases that needed to be fixed, but is consistent with respect to
behavior of the different scenarios in which the circular wait
might happen.
- Make send_row_on_empty_set() return FALSE when simplify_cond() has found out
that HAVING is always FALSE
re-committing to put the fix into 5.0 and 5.1
Mixing aggregate functions and non-grouping columns is not allowed in the
ONLY_FULL_GROUP_BY mode. However in some cases the error wasn't thrown because
of insufficient check.
In order to check more thoroughly the new algorithm employs a list of outer
fields used in a sum function and a SELECT_LEX::full_group_by_flag.
Each non-outer field checked to find out whether it's aggregated or not and
the current select is marked accordingly.
All outer fields that are used under an aggregate function are added to the
Item_sum::outer_fields list and later checked by the Item_sum::check_sum_func
function.
The name resolution for correlated subqueries and HAVING clauses
failed to distinguish which of two was being performed when there
was a reference to an outer aliased field.
Fixed by adding the condition that HAVING clause name resulotion
is being performed.
The bug is a regression introduced by the fix for bug30596. The problem
was that in cases when groups in GROUP BY correspond to only one row,
and there is ORDER BY, the GROUP BY was removed and the ORDER BY
rewritten to ORDER BY <group_by_columns> without checking if the
columns in GROUP BY and ORDER BY are compatible. This led to
incorrect ordering of the result set as it was sorted using the
GROUP BY columns. Additionaly, the code discarded ASC/DESC modifiers
from ORDER BY even if its columns were compatible with the GROUP BY
ones.
This patch fixes the regression by checking if ORDER BY columns form a
prefix of the GROUP BY ones, and rewriting ORDER BY only in that case,
preserving the ASC/DESC modifiers. That check is sufficient, since the
GROUP BY columns contain a unique index.
The optimizer takes different execution paths during EXPLAIN than SELECT,
this fix relates only to EXPLAIN, hence no behavior changes.
The test of sort keys for ORDER BY was prohibited from considering keys
that were mentioned in IGNORE KEYS FOR ORDER BY. This led to two
inconsistencies: One was that IGNORE INDEX FOR GROUP BY and
IGNORE INDEX FOR ORDER BY gave apparently different EXPLAINs; the latter
erroneously claimed to do filesort. The second inconsistency
is that the test of sort keys is called twice, finding a sort key the first
time but not the second time, leading to the mentioned filesort.
Fixed by making the test of sort keys consider all enabled
keys on the table. This test rejects keys that are not covering, and for
covering keys the hint should be ignored anyway.
and strategy (explain)
The fix for WL3527 adds tests that test if the index usage hints
combinations don't cause syntax errors.
The EXPLAIN for one of these tests can be affected by the size of the
rowid on the disk (affected by the presence of large file support).
Fixed to avoid the platform dependent test result by removing the
irrelevant columns from the EXPLAIN result.
The optimization that uses a unique index to remove GROUP BY did not
ensure that the index was actually used, thus violating the ORDER BY
that is implied by GROUP BY.
Fixed by replacing GROUP BY with ORDER BY if the GROUP BY clause contains
a unique index over non-nullable field(s). In case GROUP BY ... ORDER BY
null is used, GROUP BY is simply removed.
The optimization that uses a unique index to remove GROUP BY, did not
ensure that the index was actually used, thus violating the ORDER BY
that is impled by GROUP BY.
Fixed by replacing GROUP BY with ORDER BY if the GROUP BY clause contains
a unique index. In case GROUP BY ... ORDER BY null is used, GROUP BY is
simply removed.
SELECT statement itself returns empty.
As a result of this bug 'SELECT AGGREGATE_FUNCTION(fld) ... GROUP BY'
can return one row instead of an empty result set.
When GROUP BY only has fields of constant tables
(with a single row), the optimizer deletes the group_list.
After that we lose the information about whether we had an
GROUP BY statement. Though it's important
as SELECT min(x) from empty_table; and
SELECT min(x) from empty_table GROUP BY y; have to return
different results - the first query should return one row,
second - an empty result set.
So here we add the 'group_optimized_away' flag to remember this case
when GROUP BY exists in the query and is removed
by the optimizer, and check this flag in end_send_group()
On many architectures, e.g. 68000, x86, the double registers have higher precision
than the IEEE standard prescribes. When compiled with flags -O and higher, some double's
go into registers and therefore have higher precision. In one test case the cost
information of the best and second-best key were close enough to be influenced by this
effect, causing a failed test in distribution builds.
Fixed by removing some rows from the table in question so that cost information is not
influenced by decimals beyond standard definition of double.
When fields are inserted instead of * in the select list they were not marked
for check for the ONLY_FULL_GROUP_BY mode.
The Field_iterator_table::create_item() function now marks newly created
items for check when in the ONLY_FULL_GROUP_BY mode.
The setup_wild() and the insert_fields() functions now maintain the
cur_pos_in_select_list counter for the ONLY_FULL_GROUP_BY mode.
can be specified
Currently MySQL allows one to specify what indexes to ignore during
join optimization. The scope of the current USE/FORCE/IGNORE INDEX
statement is only the FROM clause, while all other clauses are not
affected.
However, in certain cases, the optimizer
may incorrectly choose an index for sorting and/or grouping, and
produce an inefficient query plan.
This task provides the means to specify what indexes are
ignored/used for what operation in a more fine-grained manner, thus
making it possible to manually force a better plan. We do this
by extending the current IGNORE/USE/FORCE INDEX syntax to:
IGNORE/USE/FORCE INDEX [FOR {JOIN | ORDER | GROUP BY}]
so that:
- if no FOR is specified, the index hint will apply everywhere.
- if MySQL is started with the compatibility option --old_mode then
an index hint without a FOR clause works as in 5.0 (i.e, the
index will only be ignored for JOINs, but can still be used to
compute ORDER BY).
See the WL#3527 for further details.
Currently in the ONLY_FULL_GROUP_BY mode no hidden fields are allowed in the
select list. To ensure this each expression in the select list is checked
to be a constant, an aggregate function or to occur in the GROUP BY list.
The last two requirements are wrong and doesn't allow valid expressions like
"MAX(b) - MIN(b)" or "a + 1" in a query with grouping by a.
The correct check implemented by the patch will ensure that:
any field reference in the [sub]expressions of the select list
is under an aggregate function or
is mentioned as member of the group list or
is an outer reference or
is part of the select list element that coincide with a grouping element.
The Item_field objects now can contain the position of the select list
expression which they belong to. The position is saved during the
field's Item_field::fix_fields() call.
The non_agg_fields list for non-aggregated fields is added to the SELECT_LEX
class. The SELECT_LEX::cur_pos_in_select_list now contains the position in the
select list of the expression being currently fixed.
When resolving unqualified name references MySQL was not
checking what is the item type for the reference. Thus
e.g a string literal item that has by convention a name
equal to its string value will also work as a reference to
a SELECT list item or a table field.
Fixed by allowing only Item_ref or Item_field to referenced by
(unqualified) name.
Currently SQL_BIG_RESULT is checked only at compile time.
However, additional optimizations may take place after
this check that change the sort method from 'filesort'
to sorting via index. As a result the actual plan
executed is not the one specified by the SQL_BIG_RESULT
hint. Similarly, there is no such test when executing
EXPLAIN, resulting in incorrect output.
The patch corrects the problem by testing for
SQL_BIG_RESULT both during the explain and execution
phases.
optimizer does not honor IGNORE INDEX
- Allow an index to be used for sorting the table
instead of filesort only if it is not disabled by
IGNORE INDEX.
When making a place to store field values at the start of each group
the real item (not the reference) must be used when deciding which column
to copy.
