Commit graph

65373 commits

Author SHA1 Message Date
Monty
14b1c8c80d After merge and bug fixes
- Fixed compiler warnings
- Removed have_debug.inc from innochecksum_3
- Fixed race condition in innodb_buffer_pool_load
- Fixed merge issue in innodb-bad-key-change.test
- Fixed missing array allocation that could cause
  function_defaults_notembedded to fail
- Fixed thread_cache_size_func
2016-10-05 01:11:08 +03:00
Monty
af7490f95d Remove end . from error messages to get them consistent
Fixed a few failing tests
2016-10-05 01:11:08 +03:00
Monty
c1125c3218 Fixed compiler warnings and failing tests 2016-10-05 01:11:08 +03:00
Monty
7b96416f3c Use sql_mode_t for sql_mode.
This fixed several cases where we where using just ulong for sql_mode
2016-10-05 01:11:08 +03:00
Monty
8be53a389c MDEV-6112 multiple triggers per table
This is similar to MysQL Worklog 3253, but with
a different implementation. The disk format and
SQL syntax is identical with MySQL 5.7.

Fetures supported:
- "Any" ammount of any trigger
- Supports FOLLOWS and PRECEDES to be
  able to put triggers in a certain execution order.

Implementation details:
- Class Trigger added to hold information about a trigger.
  Before this trigger information was stored in a set of lists in
  Table_triggers_list and in Table_triggers_list::bodies
- Each Trigger has a next field that poinst to the next Trigger with the
  same action and time.
- When accessing a trigger, we now always access all linked triggers
- The list are now only used to load and save trigger files.
- MySQL trigger test case (trigger_wl3253) added and we execute these
  identically.
- Even more gracefully handling of wrong trigger files than before. This
  is useful if a trigger file uses functions or syntax not provided by
  the server.
- Each trigger now has a "Created" field that shows when the trigger was
  created, with 2 decimals.

Other comments:
- Many of the changes in test files was done because of the new "Created"
  field in the trigger file. This shows up in SHOW ... TRIGGER and when
  using information_schema.trigger.
- Don't check if all memory is released if on uses --gdb;  This is needed
  to be able to get a list from safemalloc of not freed memory while
  debugging.
- Added option to trim_whitespace() to know how many prefix characters
  was skipped.
- Changed a few ulonglong sql_mode to sql_mode_t, to find some wrong usage
  of sql_mode.
2016-10-05 01:11:07 +03:00
Igor Babaev
3b314ec633 Fixed bug mdev-10933.
The bug was caused by a misplaced construct opt_with_clause
for one of the variants of CREATE ... SELECT.
2016-09-30 17:40:43 -07:00
Igor Babaev
6681a4998e Post-review addition to the fix for mdev-10868. 2016-09-30 17:40:42 -07:00
Igor Babaev
061d28207d Fixed bug mdev-10923.
The code for st_select_lex::find_table_def_in_with_clauses()
did not take into account the fact that the specs for mergeable
CTEs were cloned and were not processed by the function
With_element::check_dependencies_in_spec().
2016-09-30 17:40:42 -07:00
Igor Babaev
903f34c7a9 Fixed bug mdev-10868.
There was no implementation of the virtual method print()
for the Item_window_func class. As a result for a view
containing window function an invalid view definition could
be written in the frm file. When a query that refers to
this view was executed a syntax error was reported.
2016-09-30 17:40:40 -07:00
Igor Babaev
b91bd822fa Fixed bug mdev-10889
The bug was in the code of the recursive method
With_element::check_unrestricted_recursive. For recursive
calls of this method sel->get_with_element()->owner != owner.
2016-09-26 10:40:44 -07:00
Igor Babaev
1f1990a161 Fixed bug mdev-10884.
If a materialized derived table / view is specified by a unit
with SELECTs containing ORDER BY ... LIMIT then condition pushdown
cannot be done for these SELECTs.
If a materialized derived table / view is specified by a unit
containing global ORDER BY ... LIMIT then condition pushdown
cannot be done for this unit.
2016-09-25 17:29:10 -07:00
Igor Babaev
54efb08022 Fixed bug mdev-10881
The server missed to call check_dependencies_in_with_clauses()
when processing PREPARE ... FROM CREATE ... SELECT / INSERT ... SELECT
with WITH clause before SELECT.
2016-09-24 21:05:36 -07:00
Igor Babaev
61ab7333db Fixed bug mdev-10883.
When a prepared statement uses a CTE definition with a column list
renaming of columns of the CTE expression  must be performed
for every execution of the prepared statement.
2016-09-24 21:05:36 -07:00
Vicențiu Ciorbaru
de2175e96a Window functions can have an empty over clause
Make sure we select one window function if we have an empty over clause.
2016-09-24 15:25:00 +02:00
Vicențiu Ciorbaru
8b95e7eb44 Make sure to call Rowid_seq_cursor::next to not face infinite recursion 2016-09-24 15:12:34 +02:00
Sergei Petrunia
047963922c MDEV-9736: Window functions: multiple cursors to read filesort result
Add support for having multiple IO_CACHEs with type=READ_CACHE to share
the file they are reading from.
Each IO_CACHE keeps its own in-memory buffer. When doing a read or seek
operation on the file, it notifies other IO_CACHEs that the file position
has been changed.

Make Rowid_seq_cursor use cloned IO_CACHE when reading filesort result.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
6e4015727a Clean up nth_value
Implement nth_value correctly and add a test case for it.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
53cf265b3b Implement LEAD and LAG and NTH_VALUE functions
Refactour out (into a copy for now) the logic of Item_sum_hybrid, to
allow for multiple arguments. It does not contain the comparator
members. The result is the class Item_sum_hybrid_simple.

LEAD and LAG make use of this Item to store previous rows in a chache.
It also helps in specifying the field type. Currently LEAD/LAG do not
support default values.

NTH_VALUE behaves identical to LEAD and LAG, except that the starting
position cursor is placed on the top of the frame instead of the current
row.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
3dd3a5da0e MDEV-9935: Window functions: assertion failure with empty OVER () clause
Make window functions work with an empty over clause by forcing
a sort on the first column of the current join_tab. This is a temporary
fix until we get window functions to work with big tables.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
e992464f27 Update Frame_positional_cursor to also take an optional bound
The positional cursor now fetches rows based on the positional
cursor and an offset (if present). It will fetch rows, based on the
offset, only if the required position is not out of bounds.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
872c0b4225 Allow first/last value functions to have frame definitions
Add a test case to check this behavior.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
a2bafbabe2 Make first_value and last_value computation efficient
With clever use of partition bounds, we only need to add one row to the
items at a time. This way we remove the need to "reset" the item and run
through the full partition again.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
9a5930bcdf Implement first_value and last_value as window functions
Currently the implementation doesn't support removal, thus the
computation is performed by running over the window frame again.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
00bf18e2bf Move table record writing outside of loop
We can set values in the record buffer first and only perform one table
write call at the end. No need to write to file every time one column is
updated.

Also, remove unused method from Table_read_cursor.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
62df3119a5 Add a counter for the number of select statements using window functions
The counter is available via SHOW [GLOBAL] STATUS and will be reported
by the feedback plugin.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
5cb568786a MDEV-10669: Crash in SELECT with window function used
Make sure to call split_sum_func on all items that contain window
functions, so that all the column references are set up correctly.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
2857ff3c98 MDEV-10815: Window Function Expressions Wrong Results
Fix window function expressions such as win_func() <operator> expr.
The problem was found in 2 places.
First, when we have complex expressions containing window functions, we
can only compute their final value _after_ we have computed the window
function's values. These values must be stored within the temporary
table that we are using, before sending them off.
This is done by performing an extra copy_funcs call before the final
end_send() call.

Second, such expressions need to have their inner arguments,
changed such that the references within those arguments point to fields within
the temporary table.
Ex: sum(t.a) over (order by t.b) + sum(t.a) over (order by t.b)
Before this fix, t.a pointed to the original table's a field. In order
to compute the sum function's value correctly, it needs to point to the
copy of this field inside the temp table.
This is done by calling split_sum_func for each argument in the
expression in turn.

The win.test results have also been updated as they contained wrong
values for such a use case.
2016-09-24 15:12:34 +02:00
Vicențiu Ciorbaru
1c72441364 Frame bounds using FOLLOWING or PRECEDING can have 0 as cardinal value
This makes them behave exactly like CURRENT ROW. Standard specifies
unsigned integer, which includes the value 0.

Expand the win_min_max test to include this kind of frame definitions.
2016-09-24 15:09:39 +02:00
Sergei Petrunia
a95e384d54 MDEV-10174: Make the fix for MDEV-8989 enabled by default in 10.2
- Change the default @@optimizer_switch value
- Adjust the testcases
2016-09-24 15:27:56 +03:00
Igor Babaev
4872ec6177 Fixed bug mdev-10874.
In some cases the method Window_funcs_sort::setup() did
not build the sequence of sorting keys correctly.
2016-09-23 14:23:17 -07:00
Vladislav Vaintroub
f7a7c0c2fe MDEV-10297 Add priorization to threadpool
Also MDEV-10385 Threadpool refactoring
2016-09-22 17:01:28 +00:00
Sergei Golubchik
f32a511558 Merge branch '10.2' into bb-10.2-connector-c-integ-subm 2016-09-22 17:50:37 +02:00
Igor Babaev
48b4e33596 Allowed to use WITH clauses before SELECT in CREATE ... SELECT
and INSERT ... SELECT.
Added test cases.
2016-09-22 01:45:05 -07:00
Sergei Golubchik
59d51f0c12 Merge branch '10.2' into bb-10.2-connector-c-integ-subm 2016-09-21 12:54:56 +02:00
Sergei Golubchik
4368efe870 valgrind failures
and a couple of collaterals:
* debugging assert in my_valgrind.h
* trivial cleanup in sql_union.cc
2016-09-21 10:03:16 +02:00
Igor Babaev
10aa3936a7 Fixed bug mdev-10842.
In some cases the function compare_order_elements() erroneously
returned CMP_EQ for not equal elements.
2016-09-20 19:56:05 -07:00
Igor Babaev
3da077a6ec Merge branch '10.2' of github.com:MariaDB/server into 10.2 2016-09-19 09:54:16 -07:00
Igor Babaev
697a9d0163 Fixed a failure with --valgrind for cte_recursive.test. 2016-09-19 09:53:36 -07:00
Sergei Golubchik
f9bdc7c01a Merge branch '10.2' into bb-10.2-jan 2016-09-19 09:47:08 +02:00
Sergey Vojtovich
cb575abf76 MDEV-10296 - Multi-instance table cache
Some compilers are unhappy about mixed struct/class declarations.
2016-09-16 21:46:33 +04:00
Sergey Vojtovich
0d88b97e1e MDEV-10296 - Multi-instance table cache
Fixed type cast.
2016-09-16 19:07:59 +04:00
Sergey Vojtovich
a2b1c58c19 MDEV-10296 - Multi-instance table cache
Fixed sysvars_server_[not]embedded failure: changed type of
table_open_cache_instances from ulong to uint.

Added casts foratomic operations around tc_active_instances and
tc_contention_warning_reported: needed on some platforms.
2016-09-16 18:41:21 +04:00
Sergey Vojtovich
8613633f26 MDEV-10296 - Multi-instance table cache
Table cache instances autosizing.
2016-09-16 15:46:15 +04:00
Sergey Vojtovich
7e9ac7b8ac MDEV-10296 - Multi-instance table cache
Improve scalability by implementing multi-instance table cache.
2016-09-16 15:46:15 +04:00
Sergei Golubchik
4572dca4bd increase I_S.FILES.FILE_NANE column length
because InnoDB 5.7 tdecided to store a full path there
2016-09-16 09:54:31 +02:00
Sergei Golubchik
c8b3244703 fix some quoting in error messages
add_identifier change comes from 5.7, everything else is a follow-up
2016-09-16 09:54:30 +02:00
Monty
6c1c27ea11 Don't increment 'Empty_queries' for queries with errors. 2016-09-15 22:02:32 +03:00
Monty
7ca60dd842 Test case for fix assertion/hang in read_init_file() 2016-09-15 21:56:01 +03:00
Sergei Golubchik
b1384143e2 fix many rpl test failures
* wait_until_connected_again.inc should expect ER_CONNECTION_KILLED now
* also fix MySQL BUG#36228
2016-09-15 15:31:28 +02:00
Igor Babaev
55eb6fa5da Another attempt to fix bug mdev-10785 + cleanup for the previous attempt. 2016-09-14 11:44:41 -07:00