This patch fixes the issue with passing the DEFAULT or IGNORE values to
positional parameters for some kind of SQL statements to be executed
as prepared statements.
The main idea of the patch is to associate an actual value being passed
by the USING clause with the positional parameter represented by
the Item_param class. Such association must be performed on execution of
UPDATE statement in PS/SP mode. Other corner cases that results in
server crash is on handling CREATE TABLE when positional parameter
placed after the DEFAULT clause or CALL statement and passing either
the value DEFAULT or IGNORE as an actual value for the positional parameter.
This case is fixed by checking whether an error is set in diagnostics
area at the function pack_vcols() on return from the function pack_expression()
Subselect_single_value_engine cannot handle table value constructor used as
subquery. That's why any table value constructor TVC used as subquery is
converted into a select over derived table whose specification is TVC.
Currently the names of the columns of the derived table DT are taken from
the first element of TVC and if the k-th component of the element happens
to be a subquery the text representation of this subquery serves as the
name of the k-th column of the derived table. References of all columns of
the derived table DT compose the select list of the result of the conversion.
If a definition of a view contained a table value constructor used as a
subquery and the view was registered after this conversion had been
applied we could register an invalid view definition if the first element
of TVC contained a subquery as its component: the name of this component
was taken from the original subquery, while the name of the corresponding
column of the derived table was taken from the text representation of the
subquery produced by the function SELECT_LEX::print() and these names were
usually differ from each other.
To avoid registration of such invalid views the function SELECT_LEX::print()
now prints the original TVC instead of the select in which this TVC has
been wrapped. Now the specification of registered view looks like as if no
conversions from TVC to selects were done.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
The idea is that instead of marking all select_lex's with DISTINCT, we
only mark those that really need distinct result.
Benefits of this change:
- Temporary tables used with derived tables, UNION, IN are now smaller
as duplicates are removed already on the insert phase.
- The optimizer can now produce better plans with EQ_REF. This can be
seen from the tests where several queries does not anymore materialize
derived tables twice.
- Queries affected by 'in_predicate_conversion_threshold' where large IN
lists are converted to sub query produces better plans.
Other things:
- Removed on duplicate call to sel->init_select() in
LEX::add_primary_to_query_expression_body()
- I moved the testing of
tab->table->pos_in_table_list->is_materialized_derived()
in join_read_const_table() to the caller as it caused problems for
derived tables that could be proven to be const tables.
This also is likely to fix some bugs as if join_read_const_table()
was aborted, the table was left marked as JT_CONST, which cannot
be good. I added an ASSERT there for now that can be removed when
the code has been properly tested.
Before this patch, when calculating the cost of fetching and using a
row/key from the engine, we took into account the cost of finding a
row or key from the engine, but did not consistently take into account
index only accessed, clustered key or covered keys for all access
paths.
The cost of the WHERE clause (TIME_FOR_COMPARE) was not consistently
considered in best_access_path(). TIME_FOR_COMPARE was used in
calculation in other places, like greedy_search(), but was in some
cases (like scans) done an a different number of rows than was
accessed.
The cost calculation of row and index scans didn't take into account
the number of rows that where accessed, only the number of accepted
rows.
When using a filter, the cost of index_only_reads and cost of
accessing and disregarding 'filtered rows' where not taken into
account, which made filters cost less than there actually where.
To remedy the above, the following key & row fetch related costs
has been added:
- The cost of fetching and using a row is now split into different costs:
- key + Row fetch cost (as before) but multiplied with the variable
'optimizer_cache_cost' (default to 0.5). This allows the user to
tell the optimizer the likehood of finding the key and row in the
engine cache.
- ROW_COPY_COST, The cost copying a row from the engine to the
sql layer or creating a row from the join_cache to the record
buffer. Mostly affects table scan costs.
- ROW_LOOKUP_COST, the cost of fetching a row by rowid.
- KEY_COPY_COST the cost of finding the next key and copying it from
the engine to the SQL layer. This is used when we calculate the cost
index only reads. It makes index scans more expensive than before if
they cover a lot of rows. (main.index_merge_myisam)
- KEY_LOOKUP_COST, the cost of finding the first key in a range.
This replaces the old define IDX_LOOKUP_COST, but with a higher cost.
- KEY_NEXT_FIND_COST, the cost of finding the next key (and rowid).
when doing a index scan and comparing the rowid to the filter.
Before this cost was assumed to be 0.
All of the above constants/variables are now tuned to be somewhat in
proportion of executing complexity to each other. There is tuning
need for these in the future, but that can wait until the above are
made user variables as that will make tuning much easier.
To make the usage of the above easy, there are new (not virtual)
cost calclation functions in handler:
- ha_read_time(), like read_time(), but take optimizer_cache_cost into
account.
- ha_read_and_copy_time(), like ha_read_time() but take into account
ROW_COPY_TIME
- ha_read_and_compare_time(), like ha_read_and_copy_time() but take
TIME_FOR_COMPARE into account.
- ha_rnd_pos_time(). Read row with row id, taking ROW_COPY_COST
into account. This is used with filesort where we don't need
to execute the WHERE clause again.
- ha_keyread_time(), like keyread_time() but take
optimizer_cache_cost into account.
- ha_keyread_and_copy_time(), like ha_keyread_time(), but add
KEY_COPY_COST.
- ha_key_scan_time(), like key_scan_time() but take
optimizer_cache_cost nto account.
- ha_key_scan_and_compare_time(), like ha_key_scan_time(), but add
KEY_COPY_COST & TIME_FOR_COMPARE.
I also added some setup costs for doing different types of scans and
creating temporary tables (on disk and in memory). This encourages
the optimizer to not use these for simple 'a few row' lookups if
there are adequate key lookup strategies.
- TABLE_SCAN_SETUP_COST, cost of starting a table scan.
- INDEX_SCAN_SETUP_COST, cost of starting an index scan.
- HEAP_TEMPTABLE_CREATE_COST, cost of creating in memory
temporary table.
- DISK_TEMPTABLE_CREATE_COST, cost of creating an on disk temporary
table.
When calculating cost of fetching ranges, we had a cost of
IDX_LOOKUP_COST (0.125) for doing a key div for a new range. This is
now replaced with 'io_cost * KEY_LOOKUP_COST (1.0) *
optimizer_cache_cost', which matches the cost we use for 'ref' and
other key lookups. The effect is that the cost is now a bit higher
when we have many ranges for a key.
Allmost all calculation with TIME_FOR_COMPARE is now done in
best_access_path(). 'JOIN::read_time' now includes the full
cost for finding the rows in the table.
In the result files, many of the changes are now again close to what
they where before the "Update cost for hash and cached joins" commit,
as that commit didn't fix the filter cost (too complex to do
everything in one commit).
The above changes showed a lot of a lot of inconsistencies in
optimizer cost calculation. The main objective with the other changes
was to do calculation as similar (and accurate) as possible and to make
different plans more comparable.
Detailed list of changes:
- Calculate index_only_cost consistently and correctly for all scan
and ref accesses. The row fetch_cost and index_only_cost now
takes into account clustered keys, covered keys and index
only accesses.
- cost_for_index_read now returns both full cost and index_only_cost
- Fixed cost calculation of get_sweep_read_cost() to match other
similar costs. This is bases on the assumption that data is more
often stored on SSD than a hard disk.
- Replaced constant 2.0 with new define TABLE_SCAN_SETUP_COST.
- Some scan cost estimates did not take into account
TIME_FOR_COMPARE. Now all scan costs takes this into
account. (main.show_explain)
- Added session variable optimizer_cache_hit_ratio (default 50%). By
adjusting this on can reduce or increase the cost of index or direct
record lookups. The effect of the default is that key lookups is now
a bit cheaper than before. See usage of 'optimizer_cache_cost' in
handler.h.
- JOIN_TAB::scan_time() did not take into account index only scans,
which produced a wrong cost when index scan was used. Changed
JOIN_TAB:::scan_time() to take into consideration clustered and
covered keys. The values are now cached and we only have to call
this function once. Other calls are changed to use the cached
values. Function renamed to JOIN_TAB::estimate_scan_time().
- Fixed that most index cost calculations are done the same way and
more close to 'range' calculations. The cost is now lower than
before for small data sets and higher for large data sets as we take
into account how many keys are read (main.opt_trace_selectivity,
main.limit_rows_examined).
- Ensured that index_scan_cost() ==
range(scan_of_all_rows_in_table_using_one_range) +
MULTI_RANGE_READ_INFO_CONST. One effect of this is that if there
is choice of doing a full index scan and a range-index scan over
almost the whole table then index scan will be preferred (no
range-read setup cost). (innodb.innodb, main.show_explain,
main.range)
- Fixed the EQ_REF and REF takes into account clustered and covered
keys. This changes some plans to use covered or clustered indexes
as these are much cheaper. (main.subselect_mat_cost,
main.state_tables_innodb, main.limit_rows_examined)
- Rowid filter setup cost and filter compare cost now takes into
account fetching and checking the rowid (KEY_NEXT_FIND_COST).
(main.partition_pruning heap.heap_btree main.log_state)
- Added KEY_NEXT_FIND_COST to
Range_rowid_filter_cost_info::lookup_cost to account of the time
to find and check the next key value against the container
- Introduced ha_keyread_time(rows) that takes into account finding
the next row and copying the key value to 'record'
(KEY_COPY_COST).
- Introduced ha_key_scan_time() for calculating an index scan over
all rows.
- Added IDX_LOOKUP_COST to keyread_time() as a startup cost.
- Added index_only_fetch_cost() as a convenience function to
OPT_RANGE.
- keyread_time() cost is slightly reduced to prefer shorter keys.
(main.index_merge_myisam)
- All of the above caused some index_merge combinations to be
rejected because of cost (main.index_intersect). In some cases
'ref' where replaced with index_merge because of the low
cost calculation of get_sweep_read_cost().
- Some index usage moved from PRIMARY to a covering index.
(main.subselect_innodb)
- Changed cost calculation of filter to take KEY_LOOKUP_COST and
TIME_FOR_COMPARE into account. See sql_select.cc::apply_filter().
filter parameters and costs are now written to optimizer_trace.
- Don't use matchings_records_in_range() to try to estimate the number
of filtered rows for ranges. The reason is that we want to ensure
that 'range' is calculated similar to 'ref'. There is also more work
needed to calculate the selectivity when using ranges and ranges and
filtering. This causes filtering column in EXPLAIN EXTENDED to be
100.00 for some cases where range cannot use filtering.
(main.rowid_filter)
- Introduced ha_scan_time() that takes into account the CPU cost of
finding the next row and copying the row from the engine to
'record'. This causes costs of table scan to slightly increase and
some test to changed their plan from ALL to RANGE or ALL to ref.
(innodb.innodb_mysql, main.select_pkeycache)
In a few cases where scan time of very small tables have lower cost
than a ref or range, things changed from ref/range to ALL.
(main.myisam, main.func_group, main.limit_rows_examined,
main.subselect2)
- Introduced ha_scan_and_compare_time() which is like ha_scan_time()
but also adds the cost of the where clause (TIME_FOR_COMPARE).
- Added small cost for creating temporary table for
materialization. This causes some very small tables to use scan
instead of materialization.
- Added checking of the WHERE clause (TIME_FOR_COMPARE) of the
accepted rows to ROR costs in get_best_ror_intersect()
- Removed '- 0.001' from 'join->best_read' and optimize_straight_join()
to ensure that the 'Last_query_cost' status variable contains the
same value as the one that was calculated by the optimizer.
- Take avg_io_cost() into account in handler::keyread_time() and
handler::read_time(). This should have no effect as it's 1.0 by
default, except for heap that overrides these functions.
- Some 'ref_or_null' accesses changed to 'range' because of cost
adjustments (main.order_by)
- Added scan type "scan_with_join_cache" for optimizer_trace. This is
just to show in the trace what kind of scan was used.
- When using 'scan_with_join_cache' take into account number of
preceding tables (as have to restore all fields for all previous
table combination when checking the where clause)
The new cost added is:
(row_combinations * ROW_COPY_COST * number_of_cached_tables).
This increases the cost of join buffering in proportion of the
number of tables in the join buffer. One effect is that full scans
are now done earlier as the cost is then smaller.
(main.join_outer_innodb, main.greedy_optimizer)
- Removed the usage of 'worst_seeks' in cost_for_index_read as it
caused wrong plans to be created; It prefered JT_EQ_REF even if it
would be much more expensive than a full table scan. A related
issue was that worst_seeks only applied to full lookup, not to
clustered or index only lookups, which is not consistent. This
caused some plans to use index scan instead of eq_ref (main.union)
- Changed federated block size from 4096 to 1500, which is the
typical size of an IO packet.
- Added costs for reading rows to Federated. Needed as there is no
caching of rows in the federated engine.
- Added ha_innobase::rnd_pos_time() cost function.
- A lot of extra things added to optimizer trace
- More costs, especially for materialization and index_merge.
- Make lables more uniform
- Fixed a lot of minor bugs
- Added 'trace_started()' around a lot of trace blocks.
- When calculating ORDER BY with LIMIT cost for using an index
the cost did not take into account the number of row retrivals
that has to be done or the cost of comparing the rows with the
WHERE clause. The cost calculated would be just a fraction of
the real cost. Now we calculate the cost as we do for ranges
and 'ref'.
- 'Using index for group-by' is used a bit more than before as
now take into account the WHERE clause cost when comparing
with 'ref' and prefer the method with fewer row combinations.
(main.group_min_max).
Bugs fixed:
- Fixed that we don't calculate TIME_FOR_COMPARE twice for some plans,
like in optimize_straight_join() and greedy_search()
- Fixed bug in save_explain_data where we could test for the wrong
index when displaying 'Using index'. This caused some old plans to
show 'Using index'. (main.subselect_innodb, main.subselect2)
- Fixed bug in get_best_ror_intersect() where 'min_cost' was not
updated, and the cost we compared with was not the one that was
used.
- Fixed very wrong cost calculation for priority queues in
check_if_pq_applicable(). (main.order_by now correctly uses priority
queue)
- When calculating cost of EQ_REF or REF, we added the cost of
comparing the WHERE clause with the found rows, not all row
combinations. This made ref and eq_ref to be regarded way to cheap
compared to other access methods.
- FORCE INDEX cost calculation didn't take into account clustered or
covered indexes.
- JT_EQ_REF cost was estimated as avg_io_cost(), which is half the
cost of a JT_REF key. This may be true for InnoDB primary key, but
not for other unique keys or other engines. Now we use handler
function to calculate the cost, which allows us to handle
consistently clustered, covered keys and not covered keys.
- ha_start_keyread() didn't call extra_opt() if keyread was already
enabled but still changed the 'keyread' variable (which is wrong).
Fixed by not doing anything if keyread is already enabled.
- multi_range_read_info_cost() didn't take into account io_cost when
calculating the cost of ranges.
- fix_semijoin_strategies_for_picked_join_order() used the wrong
record_count when calling best_access_path() for SJ_OPT_FIRST_MATCH
and SJ_OPT_LOOSE_SCAN.
- Hash joins didn't provide correct best_cost to the upper level, which
means that the cost for hash_joins more expensive than calculated
in best_access_path (a difference of 10x * TIME_OF_COMPARE).
This is fixed in the new code thanks to that we now include
TIME_OF_COMPARE cost in 'read_time'.
Other things:
- Added some 'if (thd->trace_started())' to speed up code
- Removed not used function Cost_estimate::is_zero()
- Simplified testing of HA_POS_ERROR in get_best_ror_intersect().
(No cost changes)
- Moved ha_start_keyread() from join_read_const_table() to join_read_const()
to enable keyread for all types of JT_CONST tables.
- Made a few very short functions inline in handler.h
Notes:
- In main.rowid_filter the join order of order and lineitem is swapped.
This is because the cost of doing a range fetch of lineitem(98 rows) is
almost as big as the whole join of order,lineitem. The filtering will
also ensure that we only have to do very small key fetches of the rows
in lineitem.
- main.index_merge_myisam had a few changes where we are now using
less keys for index_merge. This is because index scans are now more
expensive than before.
- handler->optimizer_cache_cost is updated in ha_external_lock().
This ensures that it is up to date per statements.
Not an optimal solution (for locked tables), but should be ok for now.
- 'DELETE FROM t1 WHERE t1.a > 0 ORDER BY t1.a' does not take cost of
filesort into consideration when table scan is chosen.
(main.myisam_explain_non_select_all)
- perfschema.table_aggregate_global_* has changed because an update
on a table with 1 row will now use table scan instead of key lookup.
TODO in upcomming commits:
- Fix selectivity calculation for ranges with and without filtering and
when there is a ref access but scan is chosen.
For this we have to store the lowest known value for
'accepted_records' in the OPT_RANGE structure.
- Change that records_read does not include filtered rows.
- test_if_cheaper_ordering() needs to be updated to properly calculate
costs. This will fix tests like main.order_by_innodb,
main.single_delete_update
- Extend get_range_limit_read_cost() to take into considering
cost_for_index_read() if there where no quick keys. This will reduce
the computed cost for ORDER BY with LIMIT in some cases.
(main.innodb_ext_key)
- Fix that we take into account selectivity when counting the number
of rows we have to read when considering using a index table scan to
resolve ORDER BY.
- Add new calculation for rnd_pos_time() where we take into account the
benefit of reading multiple rows from the same page.
Tests with checking metadata or that cannot be run with
the view-protocol are excluded from --view-protocol.
For tests that do not allow the use of an additional connection,
the util connection is disabled with "--disable_service_connection".
Also cases with bugs for --view-protocol are disabled.
Repeating execution of a query containing the clause IN with string literals
in environment where the server variable in_predicate_conversion_threshold
is set results in server abnormal termination in case the query is run
as a Prepared Statement and conversion of charsets for string values in the
query are required.
The reason for server abnormal termination is that instances of the class
Item_string created on transforming the IN clause into subquery were created
on runtime memory root that is deallocated on finishing execution of Prepared
statement. On the other hand, references to Items placed on deallocated memory
root still exist in objects of the class table_value_constr. Subsequent running
of the same prepared statement leads to dereferencing of pointers to already
deallocated memory that could lead to undefined behaviour.
To fix the issue the values being pushed into a values list for TVC are created
by cloning their original items. This way the cloned items are allocate on
the PS memroot and as consequences no dangling pointer does more exist.
This patch fixes parsing problems concerning derived tables that use table
value constructors (TVC) with LIMIT and ORDER BY clauses of the form
((VALUES ... LIMIT ...) ORDER BY ...) as dt
The fix has to be applied only to 10.3 as 10.4 that employs a different
grammar rules has no such problems. The test cases should be merged
upstream.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
The bug caused crashes of the server when processing queries with nested
table value constructors (TVC) . It happened because the grammar rules to
parse TVC used the same global lists for both nested TVC and nesting TVC.
As a result invalid select trees were constructed for queries with nested
TVC and this led to crashes at the prepare stage.
This patch provides its own lists structures for each TVC nest level.
Besides the patch fixes a bug in the function wrap_tvc() that missed
inheritance of the SELECT_LEX::exclude_from_table_unique_test for
selects that wrapped TVCs. This inheritance is critical for specifications
of derived tables that employ nested TVCs.
Approved by dmitry.shulga@mariadb.com
used in set function
If a subselect is formed by a table value constructor (TVC) then the
following transformation is applied at the prepare stage:
VALUES (v1), ... (vn) => SELECT * FROM (VALUES (v1), ... (vn)) tvc_x.
The transformation is performed by the function wrap_tvc() that resets
THD::LEX::current select to the top level select of the result of the
transformation. After the call of wrap_tvc() in the function
Item_subselect::wrap_tvc_into_select() the field THD::LEX::current must be
reset to the same select as before the call. It was not done. As a result
if the subselect formed by a TVC was an argument of a set function then
an assertion was hit in the function Item_sum::check_sum_func().
Approved by Oleksandr Byelkin <sanja@mariadb.com>
This bug caused crashes of the server when processing queries with table
value constructors (TVC) that contained subqueries and were used itself as
subselects. For such TVCs the following transformation is applied at the
prepare stage:
VALUES (v1), ... (vn) => SELECT * FROM (VALUES (v1), ... (vn)) tvc_x.
This transformation allows to reduce the problem of evaluation of TVCs used
as subselects to the problem of evaluation of regular subselects.
The transformation is implemented in the wrap_tvc(). The code the function
to mimic the behaviour of the parser when processing the result of the
transformation. However this imitation was not free of some flaws. First
the function called the method exclude() that completely destroyed the
select tree structures below the transformed TVC. Second the function
used the procedure mysql_new_select to create st_select_lex nodes for
both wrapping select of the transformation and TVC. This also led to
constructing of invalid select tree structures.
The patch actually re-engineers the code of wrap_tvc().
Approved by Oleksandr Byelkin <sanja@mariadb.com>
on used subqueries
If a query was based on a table value constructor that contained subqueries
then EXPLAIN for such query did not contain any lines explaining the
execution plans of the subqueries.
This happened because
- no optimize() method was called for any subquery used by the table value
constructor when EXPLAIN command for the query was processed;
- EXPLAIN node created for the table value constructor itself did not
assume that some child nodes could be attached to it.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
of two table value costructors
This bug affected queries with a [NOT] IN/ANY/ALL subquery whose top level
unit contained several table value constructors.
The problem appeared because the code of the function
Item_subselect::fix_fields() that was responsible for wrapping table
value constructors encountered at the top level unit of a [NOT] IN/ANY/ALL
subquery did not take into account that the chain of the select objects
comprising the unit were not immutable.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
This patch actually fixes the bug MDEV-24675 and the bug MDEV-24618:
Assertion failure when TVC uses a row in the context expecting scalar value
The cause of these bugs is the same wrong call of the function that fixes
value expressions in the value list of a table value constructor.
The assertion failure happened when an expression in the value list is of
the row type. In this case an error message was expected, but it was not
issued because the function fix_fields_if_needed() was called for to
check fields of value expressions in a TVC instead of the function
fix_fields_if_needed_for_scalar() that would also check that the value
expressions are are of a scalar type.
The first bug happened when a table value expression used an expression
returned by single-row subselect. In this case the call of the
fix_fields_if_needed_for_scalar virtual function must be provided with
and address to which the single-row subselect has to be attached.
Test cases were added for each of the bugs.
Approved by Oleksandr Byelkin <sanja@mariadb.com>
1. Code simplification:
Item_default_value handled all these values:
a. DEFAULT(field)
b. DEFAULT
c. IGNORE
and had various conditions to distinguish (a) from (b) and from (c).
Introducing a new abstract class Item_contextually_typed_value_specification,
to handle (b) and (c), so the hierarchy now looks as follows:
Item
Item_result_field
Item_ident
Item_field
Item_default_value - DEFAULT(field)
Item_contextually_typed_value_specification
Item_default_specification - DEFAULT
Item_ignore_specification - IGNORE
2. Introducing a new virtual method is_evaluable_expression() to
determine if an Item is:
- a normal expression, so its val_xxx()/get_date() methods can be called
- or a just an expression substitute, whose value methods cannot be called.
3. Disallowing Items that are not evalualble expressions in table value
constructors.
A CTE can be defined as a table values constructor. In this case the CTE is
always materialized in a temporary table.
If the definition of the CTE contains a list of the names of the CTE
columns then the query expression that uses this CTE can refer to the CTE
columns by these names. Otherwise the names of the columns are taken from
the names of the columns in the result set of the query that specifies the
CTE.
Thus if the column names of a CTE are provided in the definition the
columns of result set should be renamed. In a general case renaming of
the columns is done in the select lists of the query specifying the CTE.
If a CTE is specified by a table value constructor then there are no such
select lists and renaming is actually done for the columns of the result
of materialization.
Now if a view is specified by a query expression that uses a CTE specified
by a table value constructor saving the column names of the CTE in the
stored view definition becomes critical: without these names the query
expression is not able to refer to the columns of the CTE.
This patch saves the given column names of CTEs in stored view definitions
that use them.
query with VALUES()
A table value constructor can be used in all contexts where a select
can be used. In particular an ORDER BY clause or a LIMIT clause or both
of them can be attached to a table value constructor to produce a new
query. Unfortunately execution of such queries was not supported.
This patch fixes the problem.
