Commit graph

172650 commits

Author SHA1 Message Date
Sachin Setiya
64bb59fce9 Galera MTR Tests: stability fixes
* remove part of galera_var_cluster_address.test that can not be tested reliably
 * reduce running time for galera_gcache_recover_manytrx.test
 * Additional wait_conditions for GAL-401.test

Signed-off-by: Sachin Setiya <sachinsetia1001@gmail.com>
2017-03-14 11:19:03 +05:30
Philip Stoev
451bf7243a Galera MTR Tests: Test for MW-313 Enforce wsrep_max_ws_rows also when binlog is enabled 2017-03-14 11:11:39 +05:30
Daniele Sciascia
9dda6cb08d MW-313 Enforce wsrep_max_ws_rows also when binlog is enabled 2017-03-14 11:11:18 +05:30
Philip Stoev
395c420f0f Galera MTR Tests: MW-305 , re-enable the test for ALTER USER 2017-03-14 07:13:22 +05:30
Philip Stoev
0a06347333 Galera MTR Tests: Test for MW-309 - Fix wsrep_max_ws_rows so that it does not affect SELECT queries 2017-03-14 07:12:50 +05:30
Philip Stoev
5d9c747193 Galera MTR tests: Update galera_defaults.result for GAL-360 2017-03-14 07:12:16 +05:30
Sachin Setiya
16e683fdad Galera MTR Tests: Tests for GAL-419 Respect safe_to_bootstrap flag also with
gcomm://

Signed-off-by: Sachin Setiya <sachinsetia1001@gmail.com>
2017-03-14 07:11:17 +05:30
Philip Stoev
108fd77486 Galera MTR Tests: GAL-405 Initial implementation of GCache recovery on startup. 2017-03-13 15:45:49 +05:30
Sachin Setiya
9be994ba69 MW-309 Fix wsrep_max_ws_rows so that it does not affect queries
Option wsrep_max_ws_rows is intended to limit the maximum number of rows
in a writeset. To enforce this limit, we increment THD::wsrep_affected_rows
on every INSERT, UPDATE or DELETE. The problem is that we do so even on
insertion to internal temporary tables used for SELECTs and such.
THD::wsrep_affected_rows is now incremented only for rows that are actually
replicated.

Signed-off-by: Sachin Setiya <sachinsetia1001@gmail.com>
2017-03-13 05:30:02 +05:30
Sachin Setiya
4573924f7d Galera MTR Tests: MW-308 , MW-307, GCF-992
* a dedicated test for wsrep_retry_autocommit
* some galera_toi_* tests were only passing because wsrep_retry_autocommit
  was in effect. The tests were changed to do not use autocommit
* higher timeout values in galera_2nodes.cnf , galera_3nodes.cnf

# Conflicts:
#	mysql-test/suite/galera/galera_2nodes.cnf
#	mysql-test/suite/galera/r/galera_defaults.result
#	mysql-test/suite/galera_3nodes/galera_3nodes.cnf
2017-03-12 23:00:20 +05:30
Philip Stoev
c2eaae268d Galera MTR Tests: GCF-981 - galera_bf_abort is non deterministic 2017-03-12 17:26:42 +05:30
Philip Stoev
0e105bc1f2 Galera MTR Tests: Test for GCF-942 - safe_to_bootstrap flag in grastate.dat 2017-03-12 15:08:08 +05:30
Philip Stoev
15298689cb Bump WSREP_PATCH_VERSION to 17 2017-03-12 14:24:57 +05:30
Sachin Setiya
86ec6c221a MW-267: followup to the original pull request, removed unnecessary cast.
Signed-off-by: Sachin Setiya <sachinsetia1001@gmail.com>
2017-03-12 14:24:42 +05:30
Alexey Yurchenko
3045b60f0f GAL-401: MTR test for the fix. 2017-03-12 13:45:40 +05:30
Nirbhay Choubey
5fc1ba604e Fix for post-merge build failure. 2017-01-13 13:57:17 -05:00
Nirbhay Choubey
ee8b5c305a Merge tag 'mariadb-10.0.29' into 10.0-galera 2017-01-13 13:53:59 -05:00
Vicențiu Ciorbaru
4f53384678 Merge branch 'bb-10.0-vicentiu' into 10.0
Extra merge commit due to intermediate commits pushed to 10.0 while
merge was done.
2017-01-12 03:37:35 +02:00
Vicențiu Ciorbaru
1c5ca7c183 Merge branch '5.5' into 10.0 2017-01-12 03:37:13 +02:00
Nirbhay Choubey
ab93a4d4df MDEV-11685: sql_mode can't be set with non-ascii connection charset
The supplied sql_mode(s) should be converted to ASCII first,
before comparing it with the sql_mode set.
2017-01-11 09:08:17 -05:00
Marko Mäkelä
833fda8f1a InnoDB: Enable UNIV_DEBUG_VALGRIND for cmake -DWITH_VALGRIND
The symbol HAVE_VALGRIND_MEMCHECK_H was never defined.
Instead, the symbol HAVE_VALGRIND should have been used.
2017-01-11 14:13:30 +02:00
Sergei Golubchik
c1a23cd4e5 MDEV-11676 Starting service with mysqld_safe_helper fails in SELINUX "enforcing" mode
compile, and install selinux policy for mysqld_safe_helper on centos6.
the policy was created as described in https://mariadb.com/kb/en/mariadb/what-to-do-if-mariadb-doesnt-start/#other-selinux-changes
2017-01-11 11:55:11 +01:00
Elena Stepanova
f516db3500 Updated list of unstable tests for 10.0.29 release 2017-01-11 04:45:47 +02:00
iangilfillan
9a4bc0d098 Update mysql_secure_installation man page 2017-01-10 16:05:51 +02:00
Sergei Golubchik
6ad3dd6054 mysqld_safe: don't close stdout if set -x 2017-01-10 14:19:11 +01:00
Marko Mäkelä
78e6fafcaa Fix an innodb_plugin leak noted in MDEV-11686
buf_flush_init_flush_rbt() was called too early in MariaDB server 10.0,
10.1, MySQL 5.5 and MySQL 5.6. The memory leak has been fixed in
the XtraDB storage engine and in MySQL 5.7.

As a result, when the server is started to initialize new data files,
the buf_pool->flush_rbt will be created unnecessarily and then leaked.
This memory leak was noticed in MariaDB server 10.1 when running the
test encryption.innodb_first_page.
2017-01-10 14:28:05 +02:00
Vicențiu Ciorbaru
4799af0925 Fix unit test after merge from mysql 5.5.35 perfschema
The problem in MariaDB is introduced by this merge commit:
c33db2cdc0

The merge comes from mysql and the original author comes from this
commit from MySQL:
------------------------------------------------
    commit 160b823d146288d66638e4a740d6d2da72f9a689
    Author: Marc Alff <marc.alff@oracle.com>
    Date:   Tue Aug 30 12:14:07 2016 +0200

    Bug#22551677 SIGNAL 11 IN LF_PINBOX_PUT_PINS

    Backport to 5.6
------------------------------------------------

The breaking change is in start_socket_wait_v1 where instead of using
m_thread_owner, we make use of my_pthread_getspecific_ptr to fetch a
thread local storage value. Unfortunately this invalidates the
"m_thread_owner" member when a socket is created. The internals of the
socket structure have m_thread_owner set to NULL, but when checking for
ownership we actually look at the current thread's key store.

This seems incorrect however it is not immediately apparent why.

To not diverge from MySQL's reasoning as it is not described what the
actual problem was that this commit is trying to fix, I have adjusted the
unittest to account for this new behaviour. We destroy the current
thread in the unit test, such that the newly created socket actually has
no thread owner. The m_thread_owner is untouched in all this.
2017-01-10 14:20:43 +02:00
Vicențiu Ciorbaru
d00d46f4ce Merge remote-tracking branch 'origin/5.5' into 10.0 2017-01-10 12:34:51 +02:00
Vicențiu Ciorbaru
ecdb39a9f5 Fix problems from 5.5 merge
* Update mysqld_safe script to remove duplicated parameter --crash-script
* Make --core-file-size accept underscores as well as dashes correctly.
* Add mysqld_safe_helper to Debian and Ubuntu files.
* Update innodb minor version to 35
2017-01-10 12:33:08 +02:00
Vicențiu Ciorbaru
94e18e2987 Merge remote-tracking branch 'merge/merge-perfschema-5.6' into 10.0 2017-01-10 12:32:54 +02:00
vicentiu
682d4849ff Merge remote-tracking branch 'merge/merge-innodb-5.6' into 10.0 2017-01-10 12:20:38 +02:00
Alexey Botchkov
171e59ed47 MDEV-11548 Reproducible server crash after the 2nd ALTER TABLE ADD FOREIGN KEY IF NOT EXISTS.
Have to use 'keyname' to check the name uniqueness.
2017-01-09 23:37:42 +04:00
Alexander Barkov
3e63fde52e Adding LOAD DATA tests for MDEV-11079 and MDEV-11631
c3cf7f47f0 reverted the patch
for BUG#24487120. After merging the reverting patch from MySQL
to MariaDB the problems described in MDEV-11079 and MDEV-11631 disappeared.
Adding test cases only.
2017-01-09 14:19:02 +04:00
Monty
eed319b6fb MDEV-11317: ! is_set()' or !is_set() || (m_status == DA_OK_BULK && is_bulk_op())' fails in Diagnostics_area::set_ok_status on CREATE OR REPLACE with ARCHIVE table
Problem was with deleting non existing .frm file for a storage engine that
doesn't have .frm files (yet)

Fixed by not giving an error for non existing .frm files for storage engines
that are using discovery
Fixed also valgrind supression related to the given test case
2017-01-08 17:51:36 +02:00
Vicențiu Ciorbaru
c33db2cdc0 5.6.35 2017-01-07 15:53:37 +02:00
vicentiu
6ac84d9824 5.6.35 2017-01-07 14:24:42 +02:00
vicentiu
1a55455566 Merge remote-tracking branch 'connect/10.0' into bb-10.0-vicentiu 2017-01-06 20:25:21 +02:00
Vladislav Vaintroub
eaf6b053b8 MDEV-11087 Search path for my.ini is wrong for default installation
Add <install_root>/data/my.ini to the search path -  this my.ini location
is used since MariaDB 5.2
2017-01-06 16:25:25 +01:00
Vladislav Vaintroub
82b8741ad0 Windows : use meaningful DEFAULT_MYSQL_HOME - base directory
for the default installation.

It is now defined as "C:/Program Files/MariaDB ${MYSQL_BASE_VERSION}"
which is where installer indeed puts it by default.

It still does not cover every case -32bit installer on 64 bit Windows would put installation
root under  "C:/Program Files (x86)", but better than the path used
previously C:/MariaDB${MYSQL_BASE_VERSION}, which was never correct.
2017-01-06 16:25:25 +01:00
Vladislav Vaintroub
ae6eb7a002 MDEV-11088 Client plugins cannot be loaded by command line tools
in default installation.

Added plugin-dir to the [client] section of the generated my.ini,
so that  installed services (MSI or mysql_install_db.exe) would be able to
find plugin directory.
2017-01-06 16:25:25 +01:00
vicentiu
e9aed131ea Merge remote-tracking branch 'origin/5.5' into 10.0 2017-01-06 17:09:59 +02:00
Dmitry Lenev
e4978d26b7 MDEV-9084 Calling a stored function from a nested select from temporary table causes unpredictable behavior
Cherry-pick: f4a0af070ce49abae60040f6f32e1074309c27fb
Author: Dmitry Lenev <dmitry.lenev@oracle.com>
Date:   Mon Jul 25 16:06:52 2016 +0300

  Fix for bug  "CAN'T FIND TEMPORARY TABLE".

  Attempt to execute prepared CREATE TABLE SELECT statement which used
  temporary table in the subquery in FROM clause and stored function
  failed with unwarranted ER_NO_SUCH_TABLE error. The same happened
  when such statement was used in stored procedure and this procedure
  was re-executed.

  The problem occurred because execution of such prepared statement/its
  re-execution as part of stored procedure incorrectly set
  Query_table_list::query_tables_own_last marker, indicating the last
  table which is directly used by statement. As result temporary table
  used in the subquery was treated as indirectly used/belonging to
  prelocking list and was not pre-opened by open_temporary_tables()
  call before statement execution. Thus causing ER_NO_SUCH_TABLE errors
  since our code assumes that temporary tables need to be correctly
  pre-opened before statement execution.

  This problem became visible only in version 5.6 after patches related to
  bug 11746602/27480 "EXTEND CREATE TEMPORARY TABLES PRIVILEGE TO ALLOW
  TEMP TABLE OPERATIONS" since they have introduced pre-opening of temporary
  tables for statements.

  Incorrect setting of Query_table_list::query_tables_own_last happened
  in LEX::first_lists_tables_same() method which is called by CREATE TABLE
  SELECT implementation as part of LEX::unlink_first_table(), which temporary
  excludes table list element for table being created from the query table
  list before handling SELECT part.

  LEX::first_lists_tables_same() tries to ensure that global table list of
  the statement starts with the first table list element from the first
  statement select. To do this it moves such table list element to the head
  of the global table list. If this table happens to be last directly-used
  table for the statement, query_tables_own_last marker is pointing to it.
  Since this marker was not updated when table list element was moved we
  ended up with all tables except the first table separated by it as if
  they were not directly used by statement (i.e. belonged to prelocked
  tables list).

  This fix changes code of LEX::first_lists_tables_same() to update
  query_tables_own_last marker in cases when it points to the table
  being moved. It is set to the table which precedes table being moved
  in this case.
2017-01-06 10:46:21 +01:00
Kristian Nielsen
43378f367c MDEV-10271: Stopped SQL slave thread doesn't print a message to error log like IO thread does
Make the slave SQL thread always output to the error log the message "Slave
SQL thread exiting, replication stopped in ..." whenever it previously
outputted "Slave SQL thread initialized, starting replication ...".

Before this patch, it was somewhat inconsistent in which cases the message
would be output and in which not, depending on the exact time and cause of
the condition that caused the SQL thread to stop.
2017-01-06 10:46:20 +01:00
Elena Stepanova
670b85804c Replication tests fail on valgrind due to waiting-related timeouts
MTR raises default wait_for_pos_timeout from 300 to 1500 when tests
are run with valgrind. The same needs to be done for other
replication-related waits.

The change should fix one of failures mentioned in MDEV-10653
(rpl.rpl_parallel fails in buildbot with timeout), the one
on the valgrind builder; but not all of them
2017-01-06 00:09:50 +02:00
Elena Stepanova
b2b6cf492e MDEV-10988 Sphinx test suite refuses to run silently
Add diagnostics output if any Sphinx components aren't found
2017-01-06 00:07:43 +02:00
Igor Babaev
ae1b3d1991 Fixed bug mdev-10705.
The fix for bug mdev-5104 did not take into account that
for any call of setup_order the size of ref_array must
be big enough. This patch fixes this problem.
2017-01-05 13:54:31 -08:00
Marko Mäkelä
f0c19b6a57 MDEV-11730 Memory leak in innodb.innodb_corrupt_bit
Memory was leaked when ALTER TABLE is attempted on a table
that contains corrupted indexes.
The memory leak was reported by AddressSanitizer for the test
innodb.innodb_corrupt_bit. The leak was introduced into
MariaDB Server 10.0.26, 10.1.15, 10.2.1 by the following:

commit c081c978a2
Merge: 1d21b22155 a482e76e65
Author: Sergei Golubchik <serg@mariadb.org>
Date:   Tue Jun 21 14:11:02 2016 +0200

   Merge branch '5.5' into bb-10.0
2017-01-05 20:30:51 +02:00
Elena Stepanova
9e528d4fde MDEV-11727 Sequences of tests fail with valgrind warnings in buildbot
The warning is "blocks are still reachable in loss record",
happens in malloc / _dl_close_worker. Suppression added
2017-01-05 17:38:55 +02:00
Elena Stepanova
5302ef2c95 MDEV-11700 funcs_2.innodb_charset fails in buldbot on valgrind builder with timeout
When the test is run as a part of the suite with valgrind,
only allow it to be executed if --big-test is set.
If the test is run by specifying its name explicitly, it
will still be executed, even with valgrind without big-test,
MTR has special logic for that
2017-01-05 17:33:48 +02:00
Elena Stepanova
f1ee011a6c MDEV-11722 main.join_cache fails in buildbot on very slow builders
The guilty part of the test checks for performance degradation on
a query with numerous joins on an empty table. The test expects
the query to take less than 1 second, and fails if it is not so
(which can happen on very slow builders).

The solution is to add more JOINs to the query. On a fixed server,
it should not have any noticeable impact on the query execution,
while on the unfixed version the query would take several times
longer (e.g. 6.5 sec vs 1.5 sec). Thus, we can increase the margin
for the error, and make the test fail when the query takes longer
than 5 seconds.
2017-01-05 17:32:11 +02:00