Commit graph

32575 commits

Author SHA1 Message Date
Igor Babaev
c7db46a242 Merge 5.3-5.5 2013-10-13 13:43:29 -07:00
Igor Babaev
78b580b779 Fixed bug mdev-5132.
Objects of the classes Item_func_isnull and Item_func_isnotnull
must have the flag sargable set to TRUE.
Set the value of the flag sargable only in constructors of the 
classes inherited from Item_int_func.
2013-10-11 23:24:57 -07:00
Igor Babaev
092a238864 Fixed a problem of the patch for mdev-5105 that caused valgrind complains. 2013-10-11 12:50:30 -07:00
Sergey Petrunya
3e8f7c8516 MDEV-5122: "Commands out of sync", "Malformed packet" or client hang up...
- When INSERT catches an error, it should not attempt to send parts of 
  EXPLAIN output.
2013-10-11 19:27:53 +04:00
unknown
3cc3bcfcc5 MDEV-5107:Left Join Yields All Nulls Instead of Appropriate Matches
MDEV-5034:Wrong result on LEFT JOIN with a SELECT SQ or a merge view, UNION in IN subquery

Make reset null_row same as it was set in evaluate_null_complemented_join_record().

The problem was that view firlds detect null_row by not-yet-reset table.
2013-10-11 15:57:19 +03:00
unknown
d61cffa6b1 MDEV-5130: More precise binlog position reporting for IO thread when reconnecting with GTID
Make IO thread check for end of event group, so that upon
disconnect at the end of an event group it can report the last
read GTID as expected. Also inject a fake Rotate event at
reconnect when skipping part of an initial event group, to
give SQL thread the correct Read_Master_Log_Pos.

Reported by Pavel Ivanov.
2013-10-11 11:21:18 +02:00
Igor Babaev
7c87385e30 Fixed bug mdev-5105.
The bug caused a memory overwrite in the function update_ref_and_keys()
It happened due to a wrong value of SELECT_LEX::cond_count. This value
historically was calculated by the fix_fields method. Now the logic of
calling this method became too complicated and, as a result, this value
is calculated not always correctly.
The patch changes the way how and when  the values of SELECT_LEX::cond_count
and of SELECT_LEX::between_count are calculated. The new code does it just at
the beginning of update_ref_and_keys().
2013-10-10 10:08:26 -07:00
Sergey Petrunya
094f70fd19 MDEV-5106: Server crashes in Explain_union::print_explain on ER_TOO_BIG_SELECT
- Don't save UNION's EXPLAIN data if optimization failed with 
  an error. We could end up saving incomplete plan, which will 
  cause a crash when we attempt to print it.
2013-10-10 20:30:32 +04:00
Alexey Botchkov
508d40fbdb MDEV-4788 check mysql-5.5 changes in spatial.cc.
Additional patch for the 5.5.
2013-10-10 14:20:35 +05:00
Sergey Petrunya
7e919c52a5 MDEV-3798: EXPLAIN UPDATE/DELETE
- Produce correct `key_len` when type=index.
2013-10-09 17:15:34 +04:00
Alexey Botchkov
ae82601035 MDEV-3856 Import of a large polygon fails/hangs.
The Gis_point::init_from_wkt called the String::realloc(),
        and this call is quite slow in the DEBUG mode. Which makes
        loading the huge polygon hang forever.
        Fixed by using the String::realloc(size, inc_size) version instead
        as it's done for other spatial features.
2013-10-09 17:30:50 +05:00
Sergey Petrunya
3c6ac6694d MDEV-3798: EXPLAIN UPDATE/DELETE
- Produce correct #rows for ORDER BY ... LIMIT N queries that take advantage of 
  ordered index read to read only N rows.
2013-10-09 13:07:46 +04:00
Sergey Petrunya
161d687594 MDEV-3798: EXPLAIN UPDATE/DELETE
- Generate correct contents of `Extra` column for UPDATEs/DELETEs that use quick selects
- UPDATEs with used_key_is_modified=true will show "Using buffer"
2013-10-09 09:40:33 +04:00
Alexander Barkov
1bcd2bebc6 MDEV-4425 Regexp enhancements
Do not pass PCRE_UCP flag for binary data.
This makes bytes 0x80..FF not to belong to 
generic character classes \d (digit) and \w (word character).

SELECT 0xFF RLIKE '\\w';
 -> 0

Note, this change does not affect non-binary data,
which is still examined with the PCRE_UCP flag by default.
2013-10-08 18:25:17 +04:00
Sergey Petrunya
69e6a2bb22 MDEV-3798: EXPLAIN UPDATE/DELETE
- Update test results after last few csets
- Generate correct value for `possible_keys` column for single table UPDATE/DELETE.
2013-10-08 16:13:49 +04:00
Sergey Petrunya
fda46df620 MDEV-3798: EXPLAIN UPDATE/DELETE
- if EXPLAIN DELETE prints "Deleting all rows", it should show the 
  expected number of rows in the rows column.
2013-10-08 14:26:14 +04:00
Sergey Petrunya
98a8642fe8 MDEV-3798: EXPLAIN UPDATE/DELETE
- Add support for EXPLAIN INSERT.
2013-10-07 17:29:51 +04:00
Sergey Petrunya
69393db3d1 MDEV-3798: EXPLAIN UPDATE/DELETE
- Better EXPLAIN-saving methods for quick selects
2013-10-07 13:58:47 +04:00
Sergey Petrunya
8346a498ea MDEV-3798: EXPLAIN UPDATE/DELETE
- Run probes like MYSQL_INSERT_SELECT_START or 
  MYSQL_MULTI_DELETE_START for EXPLAIN, too.
  We should run them, because 1. EXPLAIN SELECT does it, and 
  2. MySQL also does it.
2013-10-07 13:20:22 +04:00
Sergey Petrunya
062b7bfa39 Better comments 2013-10-05 13:48:45 +04:00
Sergey Petrunya
abcf14e595 MDEV-3798: EXPLAIN UPDATE/DELETE
- Handle the case when EXPLAIN UPDATE/DELETE has pruned away all partitions.
2013-10-05 13:44:01 +04:00
Sergey Petrunya
72bc6d7364 MDEV-3798: EXPLAIN UPDATE/DELETE
- Address review feedback: more renames
2013-10-05 13:19:45 +04:00
Sergey Petrunya
6a7f8af3f9 MDEV-3798: EXPLAIN UPDATE/DELETE
- Address review feedback: rename files
2013-10-05 10:25:59 +04:00
Sergey Petrunya
fedf769f0b MDEV-3798: EXPLAIN UPDATE/DELETE
- Address review feedback: rename nearly any name used by the new EXPLAIN code.
2013-10-05 09:58:22 +04:00
Igor Babaev
ec226e553a Fixed bug mdev-5078.
For aggregated fields from views/derived tables the possible adjustment
of thd->lex->in_sum_func->max_arg_level in the function Item_field::fix_fields
must be done before we leave the function.
2013-10-04 09:51:07 -07:00
Sergey Petrunya
5e4044e92c MDEV-5093, MDEV-5094:
- Make EXPLAIN {PARTITIONS,EXTENDED} {UPDATE,DELETE} work.
2013-10-04 18:50:47 +04:00
unknown
2f4db5348a MDEV-4981: Account for queries handled by query-cache in USER_STATISTICS (and in HOST_STATISTICS)
Added collection statistics for queries resolved via query cache.
2013-10-04 08:33:09 +03:00
Alexander Barkov
c8e6cb9643 MDEV-4425 Regexp enhancements
Clean-up: moving initialization of my_string_stack_guard
and pcre_stack_guard into init_libstrings() and init_pcre().
2013-10-03 16:42:20 +04:00
Igor Babaev
00816a964f Fixed bug mdev-5028.
Apparently in a general case a short-cut for the distinct optimization
is invalid if join buffers are used to join tables after the tables whose
values are to selected.
2013-10-02 17:59:56 -07:00
Alexander Barkov
f71e253914 MDEV-4425
Adding ${CMAKE_BINARY_DIR}/pcre into search path for *.h files.
Needed for find pcre.h (which is generated from pcre.h.in) when
build directory != source directory.
2013-10-02 13:56:57 +04:00
Alexander Barkov
fcf992ba4c MDEV-4425 REGEXP enhancements
Adding pcre_stack_guard to avoid crashes in pcre_compile()
on a long recursive patterns with parenthesizes:

SELECT a RLIKE '((((...((((x)))...))))';
2013-10-02 11:58:29 +04:00
Sergey Petrunya
6519ca51dd EXPLAIN UPDATE/DELETE
- Make EXPLAIN UPDATE/DELETE work inside SPs
- Return correct error code from mysql_delete()
- EXPLAIN <multi-DELETE> will create a multi_delete object (as it 
  affects the optimization). select_result will be only used for 
  producing EXPLAIN output.
2013-10-01 17:49:03 +04:00
Alexander Barkov
59311f3ead Make PCRE use my_malloc() and my_free().
This patch also makes libstrings use my_malloc() and my_free()
in embedded server. Previously, embeddes server used malloc()
and free() in libstrings.
2013-10-01 16:39:29 +04:00
Alexander Barkov
4e433d73af MDEV-4425: Removing the remainders from the old regex library. 2013-09-30 18:48:24 +04:00
Alexander Barkov
038554ec96 Always include the local copy of pcre.h and pcreposix.h instead of the
system installed (if any).
2013-09-30 13:51:05 +04:00
Alexander Barkov
7fe391adbf Fixing compilation failure on Windows.
"PCRE_STATIC" must be defined before including pcre.h
to avoid linking errors:
- unresolved external symbol __imp_regerror
- unresolved external symbol __imp_pcre_exec
2013-09-27 16:29:05 +04:00
Alexander Barkov
285e7aa179 MDEV-4425 REGEXP enhancements 2013-09-26 18:02:17 +04:00
Sergey Petrunya
0b69c44e94 MDEV-5067: Valgrind warnings (Invalid read) in QPF_table_access::print_explain
- Query plan footprint (in new terms, "EXPLAIN structure") should always keep 
  a copy of key_name.  This is because the table might be a temporary table which
  may be already freed by the time we use query plan footprint.
2013-09-26 14:42:30 +04:00
unknown
9d83468e78 merge 5.5 -> 10.0-base 2013-09-25 21:07:06 +03:00
Sergey Petrunya
7d60030c02 MDEV-5060 Server crashes on EXPLAIN EXTENDED or EXPLAIN PARTITIONS with explain in slow_log
- If we're running explain with flags, use the same set of flags to make EXPLAIN columns 
  and contents.
2013-09-25 19:18:02 +04:00
unknown
64d6d8334f merge 5.3 -> 5.5 2013-09-25 17:16:13 +03:00
Sergey Petrunya
ac54df04d8 MDEV-5070 - EXPLAIN INSERT ... SELECT crashes on 10.0-base-explain-slowquerylog
- Add EXPLAIN output print out for INSERT/REPLACE ... SELECT
2013-09-25 17:23:22 +04:00
unknown
ec7da1561e MDEV-5039: incorrect Item_func_regex::update_used_tables()
Other fix of maybe_null problem and revert of revno: 3608 "MDEV-3873 & MDEV-3876 & MDEV-3912 : Wrong result (extra rows) with ALL subquery from a MERGE view."
2013-09-25 15:30:13 +03:00
Sergey Petrunya
28734220e7 More code cleanup 2013-09-25 16:27:47 +04:00
Sergey Petrunya
3aaeb7305b Code cleanup 2013-09-25 16:07:37 +04:00
Sergey Petrunya
c0028bd2ac - Fix incorrectly-removed piece of code JOIN::save_qpf
- update mysqld--help.result
2013-09-25 15:51:16 +04:00
Vladislav Vaintroub
11fc6b49d9 MDEV-5053 - fix cyclic dependency when building with Ninja CMake generator 2013-09-23 23:33:18 +02:00
Sergey Petrunya
ccca339459 MDEV-5047 virtual THD::~THD(): Assertion `status_var.memory_used == 0' fails
- Don't forget to delete the query plan footprint when 
  the query wasn't printed into slow query log for some reason
- ALso removed some garbage code.
2013-09-23 14:17:56 +04:00
Sergey Petrunya
d998a1635f MDEV-5045: Server crashes in QPF_query::print_explain with log_slow_verbosity='explain'
- Don't print a plan when the statement didn't produce it
- Also, add first testcase. We can't check the EXPLAIN from the slow log itself, though.
2013-09-20 17:45:24 +04:00
Sergei Golubchik
2fe0836eed 5.5.33a merge 2013-09-20 11:29:01 +02:00