The issue here is the wrong estimate of the cardinality of a partial join,
the cardinality is too high because the function table_cond_selectivity()
returns an absurd number 100 while selectivity cannot be greater than 1.
When accessing table t by outer reference t1.a via index we do not perform any
range analysis for t. Yet we see TABLE::quick_key_parts[key] and
TABLE->quick_rows[key] contain a non-zero value though these should have been
remained untouched and equal to 0.
Thus real cause of the problem is that TABLE::init does not clean the arrays
TABLE::quick_key_parts[] and TABLE::>quick_rows[].
It should have done it because the TABLE structure created for any
instance of a table can be reused for many queries.
In the function prev_record_reads where one finds the different row combinations for a
subset of partial join, it did not take into account the selectivity of tables
involved in the subset of partial join.
selectivity values fails
After having set the assertion that checks validity of selectivity values
returned by the function table_cond_selectivity() a test case from
order_by.tesst failed. The failure occurred because range optimizer could
return as an estimate of the cardinality of the ranges built for an index
a number exceeding the total number of records in the table.
The second bug is more subtle. It may happen when there are several
indexes with same prefix defined on the first joined table t accessed by
a constant ref access. In this case the range optimizer estimates the
number of accessed records of t for each usable index and these
estimates can be different. Only the first of these estimates is taken
into account when the selectivity of the ref access is calculated.
However the optimizer later can choose a different index that provides
a different estimate. The function table_condition_selectivity() could use
this estimate to discount the selectivity of the ref access. This could
lead to an selectivity value returned by this function that was greater
that 1.
This allows one to run the test suite even if any of the following
options are changed:
- character-set-server
- collation-server
- join-cache-level
- log-basename
- max-allowed-packet
- optimizer-switch
- query-cache-size and query-cache-type
- skip-name-resolve
- table-definition-cache
- table-open-cache
- Some innodb options
etc
Changes:
- Don't print out the value of system variables as one can't depend on
them to being constants.
- Don't set global variables to 'default' as the default may not
be the same as the test was started with if there was an additional
option file. Instead save original value and reset it at end of test.
- Test that depends on the latin1 character set should include
default_charset.inc or set the character set to latin1
- Test that depends on the original optimizer switch, should include
default_optimizer_switch.inc
- Test that depends on the value of a specific system variable should
set it in the test (like optimizer_use_condition_selectivity)
- Split subselect3.test into subselect3.test and subselect3.inc to
make it easier to set and reset system variables.
- Added .opt files for test that required specfic options that could
be changed by external configuration files.
- Fixed result files in rockdsb & tokudb that had not been updated for
a while.
For MDEV-15955, the fix in create_tmp_field_from_item() would cause a
compilation error. After a discussion with Alexander Barkov, the fix
was omitted and only the test case was kept.
In 10.3 and later, MDEV-15955 is fixed properly by overriding
create_tmp_field() in Item_func_user_var.
Currently for selectivity calculation we perform range analysis for a column even when we don't have any statistics(EITS).
This makes less sense but is used to catch contradiction for WHERE condition.
So the solution is to not perform range analysis for selectivity calculation for columns that do not have statistics.
materialization scan over materialization lookup
For non-mergeable semi-joins we don't store the estimates of the IN subquery in table->file->stats.records.
In the function TABLE_LIST::fetch_number_of_rows, we store the number of rows in the tables
(estimates in case of derived table/views).
Currently we don't store the estimates for non-mergeable semi-joins, which leads to a problem of selecting
materialization scan over materialization lookup.
Fixed this by storing these estimated appropriately
In file sql/opt_range.cc,when calculate_cond_selectivity_for_table() is called with optimizer_use_condition_selectivity=4 then
- thd->no_errors is set to 1
- the original value of thd->no_error is not restored to its original value
- this is causing the assertion to fail in the subsequent queries
Fixed by restoring the original value of thd->no_errors
The function Item_func_isnull::update_used_tables() must
handle the case when the predicate is over not nullable
column in a special way.
This is actually a bug of MariaDB 5.3/5.5, but it's probably
hard to demonstrate that it can cause problems there.
In the function create_key_parts_for_pseudo_indexes()
the key part structures of pseudo-indexes created for
BLOB fields were set incorrectly.
Also the key parts for long fields must be 'truncated'
up to the maximum length acceptable for key parts.
1. When min/max value is provided the null flag for it must be set to 0
in the bitmap Culumn_statistics::column_stat_nulls.
2. When the calculation of the selectivity of the range condition
over a column requires min and max values for the column then we
have to check that these values are provided.
Contains also
MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7
The failure happened because 5.7 has changed the signature of
the bool handler::primary_key_is_clustered() const
virtual function ("const" was added). InnoDB was using the old
signature which caused the function not to be used.
MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7
Fixed mutexing problem on lock_trx_handle_wait. Note that
rpl_parallel and rpl_optimistic_parallel tests still
fail.
MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan)
Reason: incorrect merge
MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan)
Reason: incorrect merge
- Fix the crash by making get_column_range_cardinality()
to handle the special case where Column_stats objects
is an all-zeros object (the question of what is the point
of having Field::read_stats point to such object remains a
mystery)
- Added a few comments. Learning the code still.
- Histogram::find_bucket() should not walk off the end of the value range.
- Address review feedback in Histogram::point_selectivity(): different handling
for zero-width buckets, and explanations.
- Fix Histogram::point_selectivity() to work in the case where the
passed value_pos=0 (or 1) and the first (or the last) bucket in the
histogram has zero value-range (i.e one value).
[Attempt #2]
- Use a new selectivity calculation formula in Histogram::point_selectivity.
The formula is different from the old one because it was developed from scratch.
it doesn't have any possible division-by-zero problems.
