Bug#46754: 'rows' field doesn't reflect partition pruning
The EXPLAIN's result in 'rows' field
was evaluated to number of rows when the table was opened
(not from the table cache) and only the partitions left
after pruning was updated with its correct number
of rows.
The evaluation of the 'rows' field was using handler::records()
which is a potentially expensive call, and ignores the partitioning
pruning.
The fix was to use the handlers stats.records after updating it
with ::info(HA_STATUS_VARIABLE) instead.
called twice in a row
Queries with nested joins could cause an infinite loop in the
server when used from SP/PS.
When flattening nested joins, simplify_joins() tracks if the
name resolution list needs to be updated by setting
fix_name_res to TRUE if the current loop iteration has done any
transformations to the join table list. The problem was that
the flag was not reset before the next loop iteration leading
to unnecessary "fixing" of the name resolution list which in
turn could lead to a loop (i.e. circularly-linked part) in that
list. This was causing problems on subsequent execution when
used together with stored procedures or prepared statements.
Fixed by making sure fix_name_res is reset on every loop
iteration.
Added code resulted in strange linking problem for embedded on Windows
Avoided by not doing this for embedded mode
It's irrelevant for embedded server anyway, --protocol will be ignored
bug #49251 (deadlock/crash with concurrent truncate table and index
statistics calculation) by backporting a solution from #54678 fixed
for 5.1 plugin and 5.5.
Make the my_compiler.h header, like my_attribute.h, part of
the distribution. This is required due to the dependency of
the former on the latter (which can undefine __attribute__).
"Access compatibility" syntax
The "wild" "DELETE FROM table_name.* ... USING ..." syntax
for multi-table DELETE statements is documented but it was
lost in the fix for the bug 30234.
The table_ident_opt_wild parser rule has been added
to restore the lost syntax.
== MYSQL_TYPE_LONGLONG
A MIN/MAX() function with a subquery as its argument could lead
to a debug assertion on debug builds or wrong data on release
ones.
The problem was a combination of the following factors:
- Item_sum_hybrid::fix_fields() might use the argument
(args[0]) to calculate 'hybrid_field_type' which was later used
to decide how the data should be sent to the client.
- Item_sum::make_field() might use the argument again to
calculate the field's type when sending result set metadata to
the client.
- The argument could be changed in between these two calls via
Item::set_arg() leading to inconsistent metadata being
reported.
Here is what was happening for the bug's test case:
1. Item_sum_hybrid::fix_fields() calculates hybrid_field_type
as MYSQL_TYPE_LONGLONG based on args[0] which is an
Item::SUBSELECT_ITEM at that time.
2. A temporary table is created to execute the
query. create_tmp_field_from_item() creates a Field_long object
according to the subselect's max_length.
3. The subselect item in Item_sum_hybrid is replaced by the
Item_field object referencing the newly created Field_long.
4. Item_sum::make_field() rightfully returns the
MYSQL_TYPE_LONG type when calculating the result set metadata.
5. When sending the actual data, Item::send() relies on the
virtual field_type() function which in our case returns
previously calculated hybrid_field_type == MYSQL_TYPE_LONGLONG.
It looks like the only solution is to never refer to the
argument's metadata after the result metadata has been
calculated in fix_fields(), since the argument itself may be
different by then. In this sense, Item_sum::make_field() should
never be used, because it may rely on the argument's metadata
and is only called after fix_fields(). The "default"
implementation in Item::make_field() should be used instead as
it relies only on field_type(), but not on the argument's type.
Fixed by removing Item_sum::make_field() so that the superclass
implementation Item::make_field() is always used.
After fix for bug 39653 the shortest available secondary index was used for
full table scan. Primary clustered key was used only if no secondary index
can be used. However, when chosen secondary index includes all fields of the
table being scanned it's better to use primary index since the amount of
data to scan is the same but the primary index is clustered.
Now the find_shortest_key function takes this into account.
Queries involving predicates of the form "const NOT BETWEEN
not_indexed_column AND indexed_column" could return wrong data
due to incorrect handling by the range optimizer.
For "c NOT BETWEEN f1 AND f2" predicates, get_mm_tree()
produces a disjunction of the SEL_ARG trees for "f1 > c" and
"f2 < c". If one of the trees is empty (i.e. one of the
arguments is not sargable) the resulting tree should be empty
as well, since the whole expression in this case is not
sargable.
The above logic is implemented in get_mm_tree() as follows. The
initial state of the resulting tree is NULL (aka empty). We
then iterate through arguments and compute the corresponding
SEL_ARG tree (either "f1 > c" or "f2 < c"). If the resulting
tree is NULL, it is simply replaced by the generated
tree. Otherwise it is replaced by a disjunction of itself and
the generated tree. The obvious flaw in this implementation is
that if the first argument is not sargable and thus produces a
NULL tree, the resulting tree will simply be replaced by the
tree for the second argument. As a result, "c NOT BETWEEN f1
AND f2" will end up as just "f2 < c".
Fixed by adding a check so that when the first argument
produces an empty tree for the NOT BETWEEN case, the loop is
aborted with an empty tree as a result. The whole idea of using
a loop for 2 arguments does not make much sense, but it was
probably used to avoid code duplication for several BETWEEN
variants.
within query
The server could crash after materializing a derived table
which requires a temporary table for grouping.
When destroying the temporary table used to execute a query for
a derived table, JOIN::destroy() did not clean up Item_fields
pointing to fields in the temporary table. This led to
dereferencing a dangling pointer when printing out the items
tree later in the outer SELECT.
The solution is an addendum to the patch for bug37362: in
addition to cleaning up items in tmp_all_fields3, do the same
for items in tmp_all_fields1, since now we have an example
where this is necessary.
dict_update_statistics_low(): Create bogus statistics for those
indexes that cannot be accessed because of the innodb_force_recovery
setting.
ha_innobase::info(): Calculate statistics for each index, even if
innodb_force_recovery is set. Fill in bogus data for those indexes
that are not accessed because of the innodb_force_recovery setting.
dict_update_statistics_low(): Create bogus statistics for those
indexes that cannot be accessed because of the innodb_force_recovery
setting.
ha_innobase::info(): Calculate statistics for each index, even if
innodb_force_recovery is set. Fill in bogus data for those indexes
that are not accessed because of the innodb_force_recovery setting.
KILL_BAD_DATA is returned
Two problems discovered with the LEAST()/GREATEST()
functions:
1. The check for a null value should happen even
after the second call to val_str() in the args. This is
important because two subsequent calls to the same
Item::val_str() may yield different results.
Fixed by checking for NULL value before dereferencing
the string result.
2. While looping over the arguments and evaluating them
the loop should stop if there was an error evaluating so far
or the statement was killed. Fixed by checking for error
and bailing out.
This patch doesn't get rid of the need to acquire the dict_sys->mutex but
reduces the need to keep the mutex locked for the duration of the query
to fsp_get_available_space_in_free_extents() from ha_innobase::info().
rb://390.
The callers should indicate that the dictionary is locked or not using
the trx->dict_operation_lock_mode == RW_X_LATCH mode. Checking explicitly
for system tables is unnecessary.
Approved by Marko on IRC.
Remove a bogus debug assertion that triggered the bug.
Add assertions precisely where records must not be delete-marked.
And a comment to clarify when the record is allowed to be delete-marked.
'CREATE TABLE IF NOT EXISTS ... SELECT' behaviour
BUG#55474, BUG#55499, BUG#55598, BUG#55616 and BUG#55777 are fixed
in this patch too.
This is the 5.1 part.
It implements:
- if the table exists, binlog two events: CREATE TABLE IF NOT EXISTS
and INSERT ... SELECT
- Insert nothing and binlog nothing on master if the existing object
is a view. It only generates a warning that table already exists.
