Commit graph

2272 commits

Author SHA1 Message Date
Monty
3cb578c001 MDEV-6152: Remove calls to current_thd while creating Item
- Part 3: Adding mem_root to push_back() and push_front()

Other things:
- Added THD as an argument to some partition functions.
- Added memory overflow checking for XML tag's in read_xml()
2015-08-27 22:21:08 +03:00
Monty
1bae0d9e56 Stage 2 of MDEV-6152:
- Added mem_root to all calls to new Item
- Added private method operator new(size_t size) to Item to ensure that
  we always use a mem_root when creating an item.

This saves use once call to current_thd per Item creation
2015-08-21 10:40:51 +04:00
Sergey Vojtovich
31e365efae MDEV-8010 - Avoid sql_alloc() in Items (Patch )
Added mandatory thd parameter to Item (and all derivative classes) constructor.
Added thd parameter to all routines that may create items.
Also removed "current_thd" from Item::Item. This reduced number of
pthread_getspecific() calls from 290 to 177 per OLTP RO transaction.
2015-08-21 10:40:39 +04:00
Nirbhay Choubey
840aefc6a3 MDEV-8590: Fix embedded build failure 2015-08-09 14:16:50 -04:00
Nirbhay Choubey
cd1a11ace3 MDEV-7205 : Galera cluster & sql_log_bin = off don't work
While sql_bin_log=1(0) is meant to control binary logging for the
current session so that the updates to do(not) get logged into the
binary log to be replicated to the async MariaDB slave. The same
should not affect galera replication.

That is, the updates should always get replicated to other galera
nodes regardless of sql_bin_log's value.

Fixed by making sure that the updates are written to binlog cache
irrespective of sql_bin_log.

Added test cases.
2015-08-08 15:04:15 -04:00
Nirbhay Choubey
91acc8b16f Merge tag 'mariadb-10.0.21' into 10.0-galera 2015-08-08 14:21:22 -04:00
Kristian Nielsen
afd59b575a Merge branch 'mdev8302-3' into 10.1 2015-08-04 15:47:30 +02:00
Kristian Nielsen
5ca061e65f Fix embedded server build 2015-08-04 15:35:04 +02:00
Kristian Nielsen
dbd205797b Merge MDEV-8302 into 10.1 2015-08-04 12:39:22 +02:00
Kristian Nielsen
9b9c5e890c MDEV-8302: Duplicate key with parallel replication
This bug is essentially another variant of MDEV-7458.

If a transaction conflict caused a deadlock kill of T2 in record_gtid()
during commit, the code would do a rollback _before_ running
rgi->unmark_start_commit(). This creates a race where following transactions
could start too early (before T2 has completed its transaction retry). This
in turn could lead to replication failure, if there was a conflict that
caused eg. duplicate key error or similar.

The fix is to remove these rollbacks (in Query_log_event::do_apply_event()
and Xid_log_event::do_apply_event(). They seem out-of-place; code in
log_event.cc generally does not roll back on error, this is handled higher
up.

In addition, because of the extreme difficulty of reproducing bugs like
MDEV-7458 and MDEV-8302, this patch adds some extra precations to try to
detect (in debug builds) or prevent (in release builds) similar bugs.
ha_rollback_trans() will now call unmark_start_commit() if needed (and
assert in debug build when a caller does rollback without unmark first).

We also add an extra check for thd->killed() so that we avoid doing
mark_start_commit() if we already have a pending deadlock kill.

And we add a missing unmark_start_commit() call in the error case, found by
the above assertion.
2015-08-04 11:40:19 +02:00
Nirbhay Choubey
ee9bdcf79a Binlog SE doesn't implement abort_transaction(), so skip warning. 2015-07-14 17:33:22 -04:00
Nirbhay Choubey
dced5146bd Merge branch '10.0-galera' into 10.1 2015-07-14 16:05:29 -04:00
Monty
9bb8b74e99 Merge branch '10.1' of github.com:MariaDB/server into 10.1
Conflicts:
	sql/item_subselect.cc

Fixed also typo in comment
2015-07-09 14:47:32 +03:00
Monty
7332af49e4 - Renaming variables so that they don't shadow others (After this patch one can compile with -Wshadow and get much fewer warnings)
- Changed ER(ER_...) to ER_THD(thd, ER_...) when thd was known or if there was many calls to current_thd in the same function.
- Changed ER(ER_..) to ER_THD_OR_DEFAULT(current_thd, ER...) in some places where current_thd is not necessary defined.
- Removing calls to current_thd when we have access to thd

Part of this is optimization (not calling current_thd when not needed),
but part is bug fixing for error condition when current_thd is not defined
(For example on startup and end of mysqld)

Notable renames done as otherwise a lot of functions would have to be changed:
- In JOIN structure renamed:
   examined_rows -> join_examined_rows
   record_count -> join_record_count
- In Field, renamed new_field() to make_new_field()

Other things:
- Added DBUG_ASSERT(thd == tmp_thd) in Item_singlerow_subselect() just to be safe.
- Removed old 'tab' prefix in JOIN_TAB::save_explain_data() and use members directly
- Added 'thd' as argument to a few functions to avoid calling current_thd.
2015-07-06 20:24:14 +03:00
Vicențiu Ciorbaru
8a1b7c9c65 [MDEV-6877] Removed unnecesary bitmap in binlog_write_row
The bitmap is now deduced from the table's read and write set.
2015-06-30 13:17:09 +03:00
Vicențiu Ciorbaru
8bd5301a1e [MDEV-6877] Add binlog writing according to columns
The logging function now writes to the binlog according
to the bitmap columns implied by the binlog_row_image variable.
2015-06-30 13:17:09 +03:00
Sergei Golubchik
55b96d2708 bump the VERSION to 10.1.6 and related changes
fix two #if MYSQL_VERSION_ID > 100105
2015-06-28 12:09:18 +02:00
Sergei Golubchik
658992699b Merge tag 'mariadb-10.0.20' into 10.1 2015-06-27 20:35:26 +02:00
Nirbhay Choubey
46024098be Merge tag 'mariadb-10.0.20' into 10.0-galera 2015-06-21 23:54:55 -04:00
Sergei Golubchik
b56ad494b4 MDEV-8287 DROP TABLE suppresses all engine errors
in ha_delete_table()
* only convert ENOENT and HA_ERR_NO_SUCH_TABLE to warnings
* only return real error codes (that is, not ENOENT and
  not HA_ERR_NO_SUCH_TABLE)
* intercept HA_ERR_ROW_IS_REFERENCED to generate backward
  compatible ER_ROW_IS_REFERENCED

in mysql_rm_table_no_locks()
* no special code to handle HA_ERR_ROW_IS_REFERENCED
* no special code to handle ENOENT and HA_ERR_NO_SUCH_TABLE
* return multi-table error ER_BAD_TABLE_ERROR <table list> only
  when there were many errors, not when there were many
  tables to drop (but only one table generated an error)
2015-06-16 23:58:40 +02:00
Sergei Golubchik
fce4ab0ab4 generalize ER_TABLE_NEEDS_UPGRADE to work for views too 2015-06-01 16:01:23 +02:00
Nirbhay Choubey
e11cad9e9d Merge tag 'mariadb-10.0.19' into 10.0-galera 2015-05-09 17:09:21 -04:00
Teemu Ollakka
00afd63984 refs made wsrep_OSU_method session variable 2015-05-08 17:41:07 -04:00
Teemu Ollakka
2df5ff2691 refs - fixed race in OSU
Make sure that thd uses the same method to begin and end
OSU operation.
2015-05-08 17:41:07 -04:00
Alexey Yurchenko
4ed9ddd30e Refs
1. factored XID-related functions to a separate wsrep_xid.cc unit.
  2. refactored them to take refrences instead of pointers where appropriate
  3. implemented wsrep_get/set_SE_position to take wsrep_uuid_t and wsrep_seqno_t instead of XID
  4. call wsrep_set_SE_position() in wsrep_sst_received() to reinitialize SE checkpoint after SST was received, avoid assert() in setting code by first checking current position.
2015-05-08 17:41:06 -04:00
Kristian Nielsen
791b0ab5db Merge 10.0 -> 10.1 2015-04-20 13:21:58 +02:00
Alexander Barkov
c6d08ecce3 Merge commit 'bc902a2bfc46add0708896c07621e3707f66d95f' into 10.1 2015-04-20 11:45:22 +04:00
Sergei Petrunia
f42064a172 Fix valgrind failure: correclty clean up handler::tracker 2015-03-25 12:20:16 +03:00
Sergei Petrunia
664ce4c507 Fix linking: move the inline functions 2015-03-24 17:35:29 +03:00
Sergei Petrunia
4106dfe89b Merge branch 'bb-10.1-explain-analyze' into 10.1 2015-03-24 16:26:42 +03:00
Sergei Petrunia
77e16ce7d6 MDEV-7648: Extra data in ANALYZE FORMAT=JSON $stmt
Switch from relying on PERFORMANCE_SCHEMA to using our
own hooks for counting the time spent reading rows from
tables.
2015-03-24 16:17:41 +03:00
Sergey Vojtovich
6bd24deab4 MDEV-7728 - Improve xid cache scalability by using lock-free hash
XID cache is now based on lock-free hash.
Also fixed lf_hash_destroy() to call alloc destructor.

Note that previous implementation had race condition when thread was accessing
XA owned by different thread. This new implementation doesn't fix it either.
2015-03-16 19:07:51 +04:00
Sergey Vojtovich
18e9c314e4 MDEV-6650 - LINT_INIT emits code in non-debug builds
Replaced all references to LINT_INIT with UNINIT_VAR and LINT_INIT_STRUCT.
Removed LINT_INIT macro.
2015-03-16 14:48:22 +04:00
Kristian Nielsen
184f718fef MDEV-7249: Performance problem in parallel replication with multi-level slaves
Parallel replication (in 10.0 / "conservative" mode) relies on binlog group
commits to group transactions that can be safely run in parallel on the
slave. The --binlog-commit-wait-count and --binlog-commit-wait-usec options
exist to increase the number of commits per group. But in case of conflicts
between transactions, this can cause unnecessary delay and reduced througput,
especially on a slave where commit order is fixed.

This patch adds a heuristics to reduce this problem. When transaction T1 goes
to commit, it will first wait for N transactions to queue up for a group
commit. However, if we detect that another transaction T2 is waiting for a row
lock held by T1, then we will skip the wait and let T1 commit immediately,
releasing locks and let T2 continue.

On a slave, this avoids the unfortunate situation where T1 is waiting for T2
to join the group commit, but T2 is waiting for T1 to release locks, causing
no work to be done for the duration of the --binlog-commit-wait-usec timeout.

(The heuristic seems reasonable on the master as well, so it is enabled for
all transactions, not just replication transactions).
2015-03-13 14:01:52 +01:00
Alexander Barkov
bc902a2bfc MDEV-7387 [PATCH] Alter table xxx CHARACTER SET utf8, CONVERT TO CHARACTER SET latin1 should fail
A contribution from Daniel Black, with minor additional enhancements.
2015-03-13 16:12:54 +04:00
Jan Lindström
8249dcaaeb MDEV-6858: enforce_storage_engine option
Merge from Percona Server enforced use of a specific storage engine
authored by Stewart Smith.

Modified to be session variable and modifiable only by SUPER. Use
similar implementation as default_storage_engine.
2015-03-12 12:17:14 +02:00
Sergei Golubchik
2db62f686e Merge branch '10.0' into 10.1 2015-03-07 13:21:02 +01:00
Sergei Golubchik
7ba2916c55 MDEV-7000 Assertion `0' failed in Protocol::end_statement() on executing DDL under innodb_fake_changes=1
correct the if() condition to match the behavior of the old code
that this if() was supposed to replace
2015-02-28 19:48:22 +01:00
Nirbhay Choubey
af651c80f7 Merge tag 'mariadb-10.0.17' into 10.0-galera
Conflicts:
	storage/innobase/include/trx0trx.h
2015-02-27 17:36:54 -05:00
Kristian Nielsen
79e9ff44d1 MDEV-7458: Deadlock in parallel replication can allow following transaction to start replicating too early
In parallel replication, don't rollback inside ha_commit_trans() in case of
error.

The rollback will be done later, but the parallel replication code needs to
run unmark_start_commit() before the rollback to properly control the
sequencing of transactions.

I did not manage to come up with a reliable automatic test case for this, but
I tested it manually.
2015-02-23 13:37:34 +01:00
Nirbhay Choubey
3c9137deeb Merge branch '5.5' into 5.5-galera
Merge banch 5.5 till revision :
fdd6c11 - MDEV-7419 Function cli_safe_read not exported

Conflicts:
	.bzrignore
	debian/dist/Debian/mariadb-galera-server-5.5.files.in
	debian/dist/Ubuntu/mariadb-galera-server-5.5.files.in
	storage/tokudb/CMakeLists.txt
	support-files/build-tags
2015-02-18 17:15:57 -05:00
Sergei Golubchik
c877610e16 wsrep: ha_abort_transaction() does NOT end the transaction 2015-02-06 16:15:17 +01:00
Sergei Golubchik
80ce0c1c9c cleanup: ha_checktype()
* error reporting was never needed
* avoid useless transformaton hton to db_type to hton
* in many cases the engine was guaranteed to exist, add asserts
* use ha_default_handlerton() instead of ha_checktype(DB_TYPE_DEFAULT)
2015-02-02 20:57:46 +01:00
Sergei Golubchik
863cfb3fa5 small cleanup, remove a useless function 2015-01-31 21:51:45 +01:00
Sergei Golubchik
4b21cd21fe Merge branch '10.0' into merge-wip 2015-01-31 21:48:47 +01:00
Nirbhay Choubey
7cda4bee0e maria-10.0.16 merge
bzr merge -r4588 maria/10.0
2015-01-26 22:54:27 -05:00
Sergei Golubchik
d9c01e4b4a 5.5 merge 2015-01-21 12:03:02 +01:00
Sergei Golubchik
e53b41a9ec cleanup 2015-01-13 19:28:03 +01:00
Nirbhay Choubey
6f4f8c5f8a MDEV-7374 : Losing connection to MySQL while running ALTER TABLE
In the special case of ALTER TABLE with >10K rows, wsrep commit
should skip if wsrep is not enabled. Added a test case.
2014-12-31 20:58:54 -05:00
Sergey Vojtovich
6dbc48ca79 MDEV-7324 - Lock-free hash for table definition cache 2014-12-28 19:46:18 +04:00