Commit graph

87708 commits

Author SHA1 Message Date
Elena Stepanova
6dda6428d2 Increase version number 2014-03-11 00:03:53 +04:00
Elena Stepanova
f62eec5223 The test had synchronization point, but did not save master position
before that, which caused indeterministic outcome
2014-03-11 00:02:22 +04:00
Michael Widenius
3b55c2fe21 Fixed MDEV-5724 "Server crashes on SQL select containing more group by and left join statements using innodb tables"
The problem was that a big record was allocated on the stack, which casued stack to run out.

Fixed by using my_safe_alloca() instead of my_alloca() when allocating records.
Now only records <= 16384 are allocated on the stack.

mysql-test/r/stack-crash.result:
  Added test case
mysql-test/t/stack-crash.test:
  Added test case
storage/maria/ma_blockrec.c:
  Use my_safe_alloca() instead of my_alloca()
storage/maria/ma_dynrec.c:
  Use my_safe_alloca() instead of my_alloca()
storage/maria/maria_def.h:
  Added MARIA_MAX_RECORD_ON_STACK
storage/maria/maria_pack.c:
  Use my_safe_alloca() instead of my_alloca()
2014-03-10 21:14:38 +02:00
Michael Widenius
b07f9f72dc Fixed MDEV-5780 "create-big fails in 10.0"
The issue was that create...trigger part of the test suite used a debug_sync point that before was never triggered (in other words, wrong meaningless test).
With the new create ... replace code the debug sync point is triggered and the test case could not handled that.

I fixed this by adding a wait and go for the debug syncpoint in the test.

Removed some compiler warnings from mysql_cond_timedwait


include/mysql/psi/mysql_thread.h:
  Removed compiler warnings
mysql-test/r/create-big.result:
  New test result
mysql-test/t/create-big.test:
  Fixed test case as create_table_select_before_check_if_exists was not before triggered by the code.
2014-03-10 14:08:12 +02:00
Sergei Golubchik
5b7cab8219 default xtradb - fixes for debian packaging 2014-03-08 12:33:51 +01:00
Sergei Golubchik
12f54d7de0 @@old_mode=zero_date_time_cast 2014-03-07 21:05:28 +01:00
Sergei Golubchik
68916bcab3 workaround for xtradb on gcc 4.1.2 RHEL5/x86, gcc atomic ops only work under -march=i686 2014-03-07 17:47:47 +01:00
Sergei Golubchik
a5fdd75980 XtraDB made the default 2014-03-07 15:21:07 +01:00
unknown
fd25574645 MDEV-5789: race between rpl_parallel_change_thread_count and slave start upon server start without --skip-slave-start
Make sure to signal the condition variable for the thread pool after
the new threads have been added to the pool.

Thanks to user nanyi607rao, who reported this bug on maria-developers@.
2014-03-07 12:08:38 +01:00
unknown
2aa619ec68 MDEV-5788: Incorrect free of rgi->deferred_events in parallel replication
When an rpl_group_info object was returned from the free list, the
rgi->deferred_events_collecting and rgi->deferred_events was not correctly
re-inited. Additionally, the rgi->deferred_events was incorrectly freed in
free_rgi(), which causes unnecessary malloc/free (or crash when re-init is not
done).

Thanks to user nanyi607rao, who reported this bug on maria-developers@.
2014-03-07 12:02:09 +01:00
Alexander Barkov
fe3c68b38d Do not use SECONDS_IN_24H in nt_servc.cc.
This constant uses my_time.h, which inclusion is not desirable in nt_servc.cc
2014-03-07 10:34:07 +04:00
Alexander Barkov
89e1715355 MDEV-5372 Make "CAST(time_expr AS DATETIME)" compatible with MySQL-5.6 (and the SQL Standard) 2014-03-07 00:21:25 +04:00
Sergey Vojtovich
b95c8ce530 MDEV-5675 - Performance: my_hash_sort_bin is called too often
Reduced number of my_hash_sort_bin() calls from 4 to 1 per query.
Reduced number of memory accesses done by my_hash_sort_bin().

Details:
- let MDL subsystem use pre-calculated hash value for hash
  inserts and deletes
- let table cache use pre-calculated MDL hash value
- MDL namespace is excluded from hash value calculation, so that
  hash value can be used by table cache as is
- hash value for MDL is calculated as resulting hash value + MDL
  namespace
- extended hash implementation to accept user defined hash function
2014-03-06 16:19:12 +04:00
unknown
ae87e63184 MDEV-4603 mysql_stmt_reset returns "commands out of sync" error 2014-03-06 11:47:22 +01:00
Sergei Golubchik
4c788b06d4 10.0-base merge 2014-03-05 23:20:10 +01:00
Sergey Petrunya
23af77d26e MDEV-5723: mysqldump -uroot unusable for multi-database operations, checks all databases
- MariaDB-5.5 part of the fix: since we can't easily fix query optimization for I_S tables, 
  run the affected-tablespaces query with semijoin=off. It happens to have a good query plan
  with that setting.
[This is a forward-port to MariaDB 10.0]
2014-03-05 02:10:06 +04:00
Sergey Petrunya
5bb978777e Merge 2014-03-05 01:57:57 +04:00
Sergei Golubchik
75124c5d2b xtradb, windows, aio: fix the bad merge 2014-03-04 22:25:34 +01:00
Michael Widenius
a692645145 Fixed bug found by Pavel Ivanov in Gtid_log_event.
- Removed double call to trans_begin() for GTID BEGIN event
- Don't set OPTION_BEGIN before calling trans_begin() as this causes extra work in trans_begin()

sql/log_event.cc:
  Removed double call to trans_begin for GTID BEGIN event.
  Don't set OPTION_BEGIN before calling trans_begin() as this causes extra work in trans_begin().
  This was done by removing parsing of "BEGIN" and instead executing trans_begin() direct.
  This is much faster, but we lost the ability logging of possible slow "BEGIN" statements. As this should never happen it can be ignored.
2014-03-04 20:37:48 +02:00
Michael Widenius
2520e8d1da Fixed timing problem in rpl_heartbeat_basic.test
mysql-test/suite/rpl/t/rpl_heartbeat_basic.test:
  Added sync_slave_with_master to remove timing problem
2014-03-04 20:32:52 +02:00
unknown
cfc2eb9dd2 MDEV-5703: [PATCH] Slave disconnects and fails to reconnect on Error_code: 1159
Patch from Tomas Matejicek

Add missing error code to is_networ_error(), to allow slave to
automatically attempt reconnection also in this case.
2014-03-04 14:37:09 +01:00
unknown
5ec49e6452 Merge MDEV-5754, MDEV-5769, and MDEV-5764 into 10.0 2014-03-04 14:32:42 +01:00
unknown
bd2a0a2389 Merge MDEV-5754, MDEV-5769, and MDEV-5764 into 10.0-base 2014-03-04 14:21:00 +01:00
unknown
b5b8210849 MDEV-5754: MySQL 5.5 slaves cannot replicate from MariaDB 10.0
The problem was when a GTID event was part of a group commit, and so contained
a commit id. The code that replaces GTID with a BEGIN event for old slaves did
not correctly handle this case.

Fix the code so that the GTID with commit id can also be properly replaced
with a BEGIN query event. The extra two bytes are in the BEGIN event replaced
with a dummy, empty time zone string.
2014-03-04 13:10:14 +01:00
Sergey Petrunya
64649e3393 MDEV-5723: mysqldump -uroot unusable for multi-database operations, checks all databases
- Make do_fill_table() use join_tab->cache_select->cond if it is present.  When 
  join_tab->cache_select->cond is present, join_tab->select_cond doesn't have any 
  conditions that are usable for I_S optimizations.
2014-03-04 15:14:40 +04:00
unknown
ec374f1e53 MDEV-5769: Slave crashes on attempt to do parallel replication from an older master
Older master has no GTID events, so such events are not available for
deciding on scheduling of event groups and so on.

With this patch, we run such events from old masters single-threaded, in the
sql driver thread.

This seems better than trying to make the parallel code handle the data from
older masters; while possible, this would require a lot of testing (as well as
possibly some extra overhead in the scheduling of events), which hardly seems
worthwhile.
2014-03-04 08:48:32 +01:00
Sergei Golubchik
016bd4fc5f MDEV-5620 CMake option to compile against an external PCRE library 2014-03-04 01:22:53 +01:00
Sergey Petrunya
24d9bf8227 MDEV-5778: Valgrind failure in innodb_ext_keys.test
- Fix valgrind failure: make test_if_order_by_key() account for extended keys 
  feature.
2014-03-04 00:41:50 +04:00
unknown
641feed481 MDEV-5764: START SLAVE UNTIL does not work with parallel replication
With parallel replication, there can be any number of events queued on
in-memory lists in the worker threads.

For normal STOP SLAVE, we want to skip executing any remaining events on those
lists and stop as quickly as possible.

However, for START SLAVE UNTIL, when the UNTIL position is reached in the SQL
driver thread, we must _not_ stop until all already queued events for the
workers have been executed - otherwise we would stop too early, before the
actual UNTIL position had been completely reached.

The code did not handle UNTIL correctly, stopping too early due to not
executing the queued events to completion. Fix this, and also implement that
an explicit STOP SLAVE in the middle (when the SQL driver thread has reached
the UNTIL position but the workers have not) _will_ cause an immediate stop.
2014-03-03 12:13:55 +01:00
Sergei Golubchik
8705d00ab6 MDEV-5748 Assertion `status_var.memory_used == 0' fails on disconnect after opening an OQGRAPH table
make sure MY_THREAD_SPECIFIC is not set for memroot that can be
transferred between threads
2014-03-02 19:01:34 +01:00
Sergei Golubchik
66883ee08d MDEV-5674 Valgrind warnings "Conditional jump or move depends on uninitialised value" in create_sort_index with small sort_buffer_size
*found_rows wasn't initialized when filesort() failed
(it didn't matter, but valgrind was unhappy)
2014-03-02 15:54:57 +01:00
Sergei Golubchik
eb9f422c43 MDEV-5667 online alter and changed field/index options
use the Alter_inplace_info::ALTER_COLUMN_OPTION flag if
field/column flags were altered.

change ha_example to use check_if_supported_inplace_alter() instead
of obsolete check_if_incompatible_data()
2014-03-02 15:02:13 +01:00
Sergei Golubchik
1b608b0b9c MDEV-5735 Selecting from SEQUENCE table with negative number hangs server 2014-03-01 13:27:04 +01:00
Sergei Golubchik
04de6ccc31 MDEV-5668 Assertion `granted_role->is_role()' fails on granting role with empty name
parser: error out on empty role names
2014-03-01 11:55:31 +01:00
Sergei Golubchik
8dce8ecfda minor cleanup 2014-03-01 10:19:42 +01:00
Sergei Golubchik
45236704e8 update InnoDB version 2014-02-28 21:46:43 +01:00
Sergei Golubchik
4b3cf4aa26 XtraDB compilation failures on Windows (again) 2014-02-28 21:04:58 +01:00
Sergei Golubchik
d85fac98bb followup for MDEV-4309: DBT-3 Q1 benchmark: Benchmark + profile a patch
don't forget to initialize ORDER::fast_field_copier_setup
2014-02-28 20:24:22 +01:00
Sergei Golubchik
41c760b121 merge 2014-02-28 10:00:31 +01:00
Igor Babaev
1a536c8daf Fixed bug mdev-5635.
After constant row substitution some field items become constant items.
The range analyzer should take into account this fact when looking for
ranges.
2014-02-27 13:54:05 -08:00
Sergei Golubchik
8d0238a6d8 MDEV-4955 discover of table non-existance on CREATE
Fix ha_table_exists() to take discovery into account correctly.
It must be able to discover both table existence (when no frm is
found) and table non-existance (when frm was found).
2014-02-27 22:43:42 +01:00
Alexander Barkov
57cdc561fc Fixing AIX compilation failires 2014-02-27 19:44:00 +04:00
Jan Lindström
11826b1bcf Enable windows builds for XtraDB. 2014-02-27 16:41:49 +02:00
Sergei Golubchik
570c1a6fef MDEV-5672 MariaDB 10.0.8 doesn't compile without perfschema
apply the upstream patch
2014-02-27 12:25:51 +01:00
Sergei Golubchik
05aba79e98 MDEV-4447 MariaDB sources should have unix-style line endings everywhere 2014-02-27 12:00:16 +01:00
Elena Stepanova
560a526bc4 MDEV-5744 OQGRAPH backing table changes not reflected in OQGRAPH table
Disable query cache for OQGRAPH
2014-02-27 14:31:39 +04:00
Sergey Petrunya
472bc477f6 Merge 2014-02-27 13:43:06 +04:00
Sergey Petrunya
d7081b2524 Update test results after fix for MDEV-5732 2014-02-27 13:40:40 +04:00
Sergei Golubchik
98c0659150 compilation error in CONNECT with ODBC 2014-02-27 08:43:54 +01:00
unknown
5fcb6d4096 MDEV-5728: BINLOG_GTID_POS(..) does not return proper error unless mysql_store_result is called
Add a check for binlog open already during the parsing, so that
BINLOG_GTID_POS() can return the error early and not delay it
until execution.
2014-02-27 08:21:41 +01:00