- 27 Jun, 2016 3 commits
-
-
Alexey Botchkov authored
Servers sets default DEFINER to empty if run with the --scip-grant-tables option. So specify the DEFINER explicitly
-
Alexey Botchkov authored
Compiler on Sparc is strict about the 'const' modifiers in function declarations and definitions. Meaning they should be exactly same.
-
Alexander Barkov authored
MDEV-10277 Redundant NOTE when inserting '0.00001 ' into a DECIMAL(2,1) column
-
- 24 Jun, 2016 1 commit
-
-
holyfoot authored
MDEV-10214: Fix segfault when using groups in PAM user mapping plugin
-
- 23 Jun, 2016 2 commits
-
-
Alexander Barkov authored
MDEV-8686 A user defined collation utf8_confusables doesn't work 25e68c5e
-
Alexander Barkov authored
The collation customization code for the UCA (Unicode Collation Alrorithm) based collations now allows to reset to and shift of characters with implicit weights. Previously reset/shift worked only for the characters with explicit DUCET weights. An attempt to use reset/shift with character with implicit weights made the server crash.
-
- 22 Jun, 2016 1 commit
-
-
Sergei Petrunia authored
Make tree_or set correct SEL_TREE::keys_map for the result.
-
- 21 Jun, 2016 1 commit
-
-
Alexander Barkov authored
-
- 19 Jun, 2016 1 commit
-
-
Monty authored
Fixed mysql-test-run failures for roles.acl_statistics
-
- 12 Jun, 2016 1 commit
-
-
Mario Weigel authored
-
- 08 Jun, 2016 2 commits
-
-
Kristian Nielsen authored
-
Kristian Nielsen authored
The check inserts a DWARF directive to tell stack unwinding that the bottom of the (co-routine) stack has been reached. Without this, stack traces may attempt to continue past the bottom of the stack. The GCC version check was incorrect, and failed to trigger for GCC version 5.[0123].
-
- 03 Jun, 2016 3 commits
-
-
Vladislav Vaintroub authored
On Windows with low precision (10-16 ms) timer used for timing queries, the statement "set @@long_query_time=0.01" can log itself a slow if it spawn more single timer tick. The fix is to filter this query out from "SELECT FROM mysql.slow_log" results
-
Sergei Petrunia authored
Make the range optimizer not call records_in_range() when it would not give any benefit.
-
Sergei Petrunia authored
-
- 02 Jun, 2016 6 commits
-
-
Sergei Petrunia authored
-
Sergei Petrunia authored
Part #2: make tree_or(tree1, tree2) to reuse tree1 for the result object for simple cases. These include key IN (c1, ... cN). The reuse was happening in old MySQL versions, but we stopped doing it in the "fair choice between range and index_merge" patch.
-
Sergei Petrunia authored
-
Sergei Petrunia authored
-
Sergei Petrunia authored
-
Jan Lindström authored
MDEV-10154 - Remove some unnecessary parentheses
-
- 01 Jun, 2016 4 commits
-
-
Sergei Petrunia authored
Post-fix #2: - Update test results - Make the optimization conditional under @@optimizer_switch flag. - The optimization is now disabled by default, so .result files are changed back to be what they were before the MDEV-8989 patch.
-
Jan Lindström authored
Add big_test and not_debug requirements to avoid timeouts.
-
Jan Lindström authored
Fix spelling: shhould -> should
-
Hyeonseok Oh authored
-
- 31 May, 2016 1 commit
-
-
Sergei Petrunia authored
A partial backport of 67f21fb3a077dedfd14b9ca720e926c55e682f93, Bug#22283790: RANGE OPTIMIZER UTILIZES TOO MUCH MEMORY WITH MANY OR CONDITIONS The backported part changes SEL_TREE::keys from being an array of MAX_KEY elements (64*8=512 bytes) to a Mem_root_array<SEL_ARG*> (32 bytes + alloc'ed array of as many elements as we need). The patch doesn't fix the "not limiting memory" part, but the memory usage is much lower with it.
-
- 30 May, 2016 4 commits
-
-
Alexander Barkov authored
utf8mb4_thai_520_w2, ucs2_thai_520_w2, utf16_thai_520_w2, utf32_thai_520_w2
-
Alexander Barkov authored
from ctype_uca.test. This is to reuse ctype_thai.inc for other Unicode character sets later - Removing separate ctype_uca.result
-
Alexander Barkov authored
and including it from t/ctype_uca.test - Deleting r/ctype_uca_th.result
-
Alexander Barkov authored
They have bytes 0x25 and 0x5F. Testing that these bytes are treated as parts of multi-byte characters rather than underscore and percent sign.
-
- 28 May, 2016 1 commit
-
-
Vladislav Vaintroub authored
Mark gssapi plugin as stable. No open bug reports, and no further work planned, thus stable is accurate
-
- 27 May, 2016 1 commit
-
-
Sergei Petrunia authored
Post-fix: add comments, remove dead-code.
-
- 26 May, 2016 4 commits
-
-
Alexander Barkov authored
- Changing strnxfrm_multiply from 8 to 4, as agreed with Pruet Boonma - Adjusting tests
-
Alexander Barkov authored
-
Hyeonseok Oh authored
-
pruet authored
-
- 23 May, 2016 1 commit
-
-
Sergei Petrunia authored
Variant #4 of the fix. Make ORDER BY optimization functions take into account multiple equalities. This is done in several places: - remove_const() checks whether we can sort the first table in the join, or we need to put rows into temp.table and then sort. - test_if_order_by_key() checks whether there are indexes that can be used to produce the required ordering - make_unireg_sortorder() constructs sort criteria for filesort.
-
- 18 May, 2016 1 commit
-
-
Alexander Barkov authored
-
- 17 May, 2016 1 commit
-
-
Sergei Petrunia authored
-
- 11 May, 2016 1 commit
-
-
Sergei Petrunia authored
When simplify_joins() converts an outer join to an inner, it should reset the value of TABLE::dep_tables. This is needed, because the function may have already set TABLE::dep_tables according to the outer join dependency.
-