Add a wait-for graph based deadlock detector to the
MDL subsystem.
Fixes bug #46272 "MySQL 5.4.4, new MDL: unnecessary deadlock" and
bug #37346 "innodb does not detect deadlock between update and
alter table".
The first bug manifested itself as an unwarranted abort of a
transaction with ER_LOCK_DEADLOCK error by a concurrent ALTER
statement, when this transaction tried to repeat use of a
table, which it has already used in a similar fashion before
ALTER started.
The second bug showed up as a deadlock between table-level
locks and InnoDB row locks, which was "detected" only after
innodb_lock_wait_timeout timeout.
A transaction would start using the table and modify a few
rows.
Then ALTER TABLE would come in, and start copying rows
into a temporary table. Eventually it would stumble on
the modified records and get blocked on a row lock.
The first transaction would try to do more updates, and get
blocked on thr_lock.c lock.
This situation of circular wait would only get resolved
by a timeout.
Both these bugs stemmed from inadequate solutions to the
problem of deadlocks occurring between different
locking subsystems.
In the first case we tried to avoid deadlocks between metadata
locking and table-level locking subsystems, when upgrading shared
metadata lock to exclusive one.
Transactions holding the shared lock on the table and waiting for
some table-level lock used to be aborted too aggressively.
We also allowed ALTER TABLE to start in presence of transactions
that modify the subject table. ALTER TABLE acquires
TL_WRITE_ALLOW_READ lock at start, and that block all writes
against the table (naturally, we don't want any writes to be lost
when switching the old and the new table). TL_WRITE_ALLOW_READ
lock, in turn, would block the started transaction on thr_lock.c
lock, should they do more updates. This, again, lead to the need
to abort such transactions.
The second bug occurred simply because we didn't have any
mechanism to detect deadlocks between the table-level locks
in thr_lock.c and row-level locks in InnoDB, other than
innodb_lock_wait_timeout.
This patch solves both these problems by moving lock conflicts
which are causing these deadlocks into the metadata locking
subsystem, thus making it possible to avoid or detect such
deadlocks inside MDL.
To do this we introduce new type-of-operation-aware metadata
locks, which allow MDL subsystem to know not only the fact that
transaction has used or is going to use some object but also what
kind of operation it has carried out or going to carry out on the
object.
This, along with the addition of a special kind of upgradable
metadata lock, allows ALTER TABLE to wait until all
transactions which has updated the table to go away.
This solves the second issue.
Another special type of upgradable metadata lock is acquired
by LOCK TABLE WRITE. This second lock type allows to solve the
first issue, since abortion of table-level locks in event of
DDL under LOCK TABLES becomes also unnecessary.
Below follows the list of incompatible changes introduced by
this patch:
- From now on, ALTER TABLE and CREATE/DROP TRIGGER SQL (i.e. those
statements that acquire TL_WRITE_ALLOW_READ lock)
wait for all transactions which has *updated* the table to
complete.
- From now on, LOCK TABLES ... WRITE, REPAIR/OPTIMIZE TABLE
(i.e. all statements which acquire TL_WRITE table-level lock) wait
for all transaction which *updated or read* from the table
to complete.
As a consequence, innodb_table_locks=0 option no longer applies
to LOCK TABLES ... WRITE.
- DROP DATABASE, DROP TABLE, RENAME TABLE no longer abort
statements or transactions which use tables being dropped or
renamed, and instead wait for these transactions to complete.
- Since LOCK TABLES WRITE now takes a special metadata lock,
not compatible with with reads or writes against the subject table
and transaction-wide, thr_lock.c deadlock avoidance algorithm
that used to ensure absence of deadlocks between LOCK TABLES
WRITE and other statements is no longer sufficient, even for
MyISAM. The wait-for graph based deadlock detector of MDL
subsystem may sometimes be necessary and is involved. This may
lead to ER_LOCK_DEADLOCK error produced for multi-statement
transactions even if these only use MyISAM:
session 1: session 2:
begin;
update t1 ... lock table t2 write, t1 write;
-- gets a lock on t2, blocks on t1
update t2 ...
(ER_LOCK_DEADLOCK)
- Finally, support of LOW_PRIORITY option for LOCK TABLES ... WRITE
was abandoned.
LOCK TABLE ... LOW_PRIORITY WRITE from now on has the same
priority as the usual LOCK TABLE ... WRITE.
SELECT HIGH PRIORITY no longer trumps LOCK TABLE ... WRITE in
the wait queue.
- We do not take upgradable metadata locks on implicitly
locked tables. So if one has, say, a view v1 that uses
table t1, and issues:
LOCK TABLE v1 WRITE;
FLUSH TABLE t1; -- (or just 'FLUSH TABLES'),
an error is produced.
In order to be able to perform DDL on a table under LOCK TABLES,
the table must be locked explicitly in the LOCK TABLES list.
The problem was that FLUSH TABLE <table_list> would block,
waiting for all tables with old versions to be removed from
the table definition cache, rather than waiting for only
the tables in <table_list>. This could happen if FLUSH TABLE
was used in combination with LOCK TABLES.
With the new MDL code, this problem is no longer repeatable.
Regression test case added to lock.test. This commit contains
no code changes.
INFILE".
Attempts to execute an INSERT statement for a MEMORY table which invoked
a trigger or called a stored function which tried to perform LOW_PRIORITY
update on the table being inserted into, resulted in debug servers aborting
due to an assertion failure. On non-debug servers such INSERTs failed with
"Can't update table t1 in stored function/trigger because it is already used
by statement which invoked this stored function/trigger" as expected.
The problem was that in the above scenario TL_WRITE_CONCURRENT_INSERT
is converted to TL_WRITE inside the thr_lock() function since the MEMORY
engine does not support concurrent inserts. This triggered an assertion
which assumed that for the same table, one thread always requests locks with
higher thr_lock_type value first. When TL_WRITE_CONCURRENT_INSERT is
upgraded to TL_WRITE after the locks have been sorted, this is no longer true.
In this case, TL_WRITE was requested after acquiring a TL_WRITE_LOW_PRIORITY
lock on the table, triggering the assert.
This fix solves the problem by adjusting this assert to take this
scenario into account.
An alternative approach to change handler::store_locks() methods for all engines
which do not support concurrent inserts in such way that
TL_WRITE_CONCURRENT_INSERT is upgraded to TL_WRITE there instead,
was considered too intrusive.
Commit on behalf of Dmitry Lenev.
Bug #48248 assert in MDL_ticket::upgrade_shared_lock_to_exclusive
The assert would happen if REPAIR TABLE was used on a table already
locked by LOCK TABLES READ. REPAIR mistakenly tried to upgrade the
read-lock to exclusive, thereby triggering the assert.
The cause of the problem was that REPAIR TABLE ignored errors
from opening and locking tables. This is by design, as REPAIR
can be used to broken tables that cannot be opened. However,
repair also ignored logical errors such as the inability to
exclusivly lock a table due to conflicting LOCK TABLES.
This patch fixes the problem by not ignoring errors from
opening and locking tables if inside LOCK TABLES mode.
In LOCK TABLES we already know that the table can be opened,
so that the failure to open must be a logical error.
Test added to repair.test.
Bug #43272 HANDLER SQL command does not work under LOCK TABLES
HANDLER commands are now explicitly disallowed in LOCK TABLES mode.
Before, HANDLER OPEN gave the misleading error message: "Table x was
not locked with LOCK TABLES". This patch changes HANDLER OPEN/READ/CLOSE
to give ER_LOCK_OR_ACTIVE_TRANSACTION "Can't execute the given command
because you have active locked tables or an active transaction" in
LOCK TABLES mode.
Test case added to lock.test.
Bug #45066 FLUSH TABLES WITH READ LOCK deadlocks against LOCK TABLE
Test coverage for combinations of LOCK TABLE READ / WRITE and
FLUSH TABLES / FLUSH TABLES WITH READ LOCK added to lock.test.
LOCK and FLUSH are executed sequentially from one connection.
----------------------------------------------------------
revno: 2617.69.2
committer: Konstantin Osipov <kostja@sun.com>
branch nick: 5.4-azalea-bugfixing
timestamp: Mon 2009-08-03 19:26:04 +0400
message:
A fix and a test case for Bug#45035 "Altering table under LOCK TABLES
results in "Error 1213 Deadlock found...".
If a user had a table locked with LOCK TABLES
for READ and for WRITE in the same connection, ALTER TABLE
could fail.
Root cause analysis:
If a connection issues
LOCK TABLE t1 write, t1 a read, t1 b read;
the new LOCK TABLES code in 6.0 (part of WL 3726) will create
the following list of TABLE_LIST objects
(thd->locked_tables_list->m_locked_tables):
{"t1" "b" tl_read_no_insert}, {"t1" "a" tl_read_no_insert},
{"t1" "t1" tl_write }
Later on, when we try to ALTER table t1, mysql_alter_table()
closes all TABLE instances and releases its thr_lock locks,
keeping only an exclusive metadata lock on t1.
But when ALTER is finished, Locked_table_list::reopen_tables()
tries to restore the original list of open and locked tables.
Before this patch, it used to do so one by one:
Open t1 b, get TL_READ_NO_INSERT lock,
Open t1 a, get TL_READ_NO_INSERT lock
Open t1, try to get TL_WRITE lock, deadlock.
The cause of the deadlock is that thr_lock.c doesn't
resolve the situation when the read list only consists
of locks taken by the same thread, followed by this very
thread trying to take a WRITE lock. Indeed, since
thr_lock_multi always gets a sorted list of locks,
WRITE locks always precede READ locks in the list
to lock.
Don't try to fix thr_lock.c deficiency, keep this
code simple.
Instead, try to take all thr_lock locks at once
in ::reopen_tables().
2617.31.12, 2617.31.15, 2617.31.15, 2617.31.16, 2617.43.1
- initial changeset that introduced the fix for
Bug#989 and follow up fixes for all test suite failures
introduced in the initial changeset.
------------------------------------------------------------
revno: 2617.31.1
committer: Davi Arnaut <Davi.Arnaut@Sun.COM>
branch nick: 4284-6.0
timestamp: Fri 2009-03-06 19:17:00 -0300
message:
Bug#989: If DROP TABLE while there's an active transaction, wrong binlog order
WL#4284: Transactional DDL locking
Currently the MySQL server does not keep metadata locks on
schema objects for the duration of a transaction, thus failing
to guarantee the integrity of the schema objects being used
during the transaction and to protect then from concurrent
DDL operations. This also poses a problem for replication as
a DDL operation might be replicated even thought there are
active transactions using the object being modified.
The solution is to defer the release of metadata locks until
a active transaction is either committed or rolled back. This
prevents other statements from modifying the table for the
entire duration of the transaction. This provides commitment
ordering for guaranteeing serializability across multiple
transactions.
- Incompatible change:
If MySQL's metadata locking system encounters a lock conflict,
the usual schema is to use the try and back-off technique to
avoid deadlocks -- this schema consists in releasing all locks
and trying to acquire them all in one go.
But in a transactional context this algorithm can't be utilized
as its not possible to release locks acquired during the course
of the transaction without breaking the transaction commitments.
To avoid deadlocks in this case, the ER_LOCK_DEADLOCK will be
returned if a lock conflict is encountered during a transaction.
Let's consider an example:
A transaction has two statements that modify table t1, then table
t2, and then commits. The first statement of the transaction will
acquire a shared metadata lock on table t1, and it will be kept
utill COMMIT to ensure serializability.
At the moment when the second statement attempts to acquire a
shared metadata lock on t2, a concurrent ALTER or DROP statement
might have locked t2 exclusively. The prescription of the current
locking protocol is that the acquirer of the shared lock backs off
-- gives up all his current locks and retries. This implies that
the entire multi-statement transaction has to be rolled back.
- Incompatible change:
FLUSH commands such as FLUSH PRIVILEGES and FLUSH TABLES WITH READ
LOCK won't cause locked tables to be implicitly unlocked anymore.
----------------------------------------------------------
revno: 2630.4.35
committer: Konstantin Osipov <konstantin@mysql.com>
branch nick: mysql-6.0-3726
timestamp: Wed 2008-06-25 16:44:00 +0400
message:
Fix a MyISAM-specific bug in the new implementation of
LOCK TABLES (WL#3726).
If more than one instance of a MyISAM table are open in the
same connection, all of them must share the same status_param.
Otherwise, unlock of a table may lead to lost records.
See also comments in thr_lock.c.
----------------------------------------------------------
revno: 2630.10.1
committer: Konstantin Osipov <konstantin@mysql.com>
branch nick: mysql-6.0-lock-tables-tidyup
timestamp: Wed 2008-06-11 15:49:58 +0400
message:
WL#3726, review fixes.
Now that we have metadata locks, we don't need to keep a crippled
TABLE instance in the table cache to indicate that a table is locked.
Remove all code that used this technique. Instead, rely on metadata
locks and use the standard open_table() and close_thread_table()
to manipulate with the table cache tables.
Removes a list of functions that have become unused (see the comment
for sql_base.cc for details).
Under LOCK TABLES, keep a TABLE_LIST instance for each table
that may be temporarily closed. For that, implement an own class for
LOCK TABLES mode, Locked_tables_list.
This is a pre-requisite patch for WL#4144.
This is not exactly a backport: there is no new
online ALTER table in Celosia, so the old alter table
code was changed to work with the new table cache API.
Backport of:
------------------------------------------------------------
revno: 2630.4.1
committer: Dmitry Lenev <dlenev@mysql.com>
branch nick: mysql-6.0-3726-w
timestamp: Fri 2008-05-23 17:54:03 +0400
message:
WL#3726 "DDL locking for all metadata objects".
After review fixes in progress.
------------------------------------------------------------
This is the first patch in series. It transforms the metadata
locking subsystem to use a dedicated module (mdl.h,cc). No
significant changes in the locking protocol.
The import passes the test suite with the exception of
deprecated/removed 6.0 features, and MERGE tables. The latter
are subject to a fix by WL#4144.
Unfortunately, the original changeset comments got lost in a merge,
thus this import has its own (largely insufficient) comments.
This patch fixes Bug#25144 "replication / binlog with view breaks".
Warning: this patch introduces an incompatible change:
Under LOCK TABLES, it's no longer possible to FLUSH a table that
was not locked for WRITE.
Under LOCK TABLES, it's no longer possible to DROP a table or
VIEW that was not locked for WRITE.
******
Backport of:
------------------------------------------------------------
revno: 2630.4.2
committer: Dmitry Lenev <dlenev@mysql.com>
branch nick: mysql-6.0-3726-w
timestamp: Sat 2008-05-24 14:03:45 +0400
message:
WL#3726 "DDL locking for all metadata objects".
After review fixes in progress.
******
Backport of:
------------------------------------------------------------
revno: 2630.4.3
committer: Dmitry Lenev <dlenev@mysql.com>
branch nick: mysql-6.0-3726-w
timestamp: Sat 2008-05-24 14:08:51 +0400
message:
WL#3726 "DDL locking for all metadata objects"
Fixed failing Windows builds by adding mdl.cc to the lists
of files needed to build server/libmysqld on Windows.
******
Backport of:
------------------------------------------------------------
revno: 2630.4.4
committer: Dmitry Lenev <dlenev@mysql.com>
branch nick: mysql-6.0-3726-w
timestamp: Sat 2008-05-24 21:57:58 +0400
message:
WL#3726 "DDL locking for all metadata objects".
Fix for assert failures in kill.test which occured when one
tried to kill ALTER TABLE statement on merge table while it
was waiting in wait_while_table_is_used() for other connections
to close this table.
These assert failures stemmed from the fact that cleanup code
in this case assumed that temporary table representing new
version of table was open with adding to THD::temporary_tables
list while code which were opening this temporary table wasn't
always fulfilling this.
This patch changes code that opens new version of table to
always do this linking in. It also streamlines cleanup process
for cases when error occurs while we have new version of table
open.
******
WL#3726 "DDL locking for all metadata objects"
Add libmysqld/mdl.cc to .bzrignore.
******
Backport of:
------------------------------------------------------------
revno: 2630.4.6
committer: Dmitry Lenev <dlenev@mysql.com>
branch nick: mysql-6.0-3726-w
timestamp: Sun 2008-05-25 00:33:22 +0400
message:
WL#3726 "DDL locking for all metadata objects".
Addition to the fix of assert failures in kill.test caused by
changes for this worklog.
Make sure we close the new table only once.
------------------------------------------------------------
revno: 2476.784.3
committer: davi@moksha.local
timestamp: Tue 2007-10-02 21:27:31 -0300
message:
Bug#25858 Some DROP TABLE under LOCK TABLES can cause deadlocks
When a client (connection) holds a lock on a table and attempts to
drop (obtain a exclusive lock) on a second table that is already
held by a second client and the second client then attempts to
drop the table that is held by the first client, leads to a
circular wait deadlock. This scenario is very similar to trying to
drop (or rename) a table while holding read locks and are
correctly forbidden.
The solution is to allow a drop table operation to continue only
if the table being dropped is write (exclusively) locked, or if
the table is temporary, or if the client is not holding any
locks. Using this scheme prevents the creation of a circular
chain in which each client is waiting for one table that the
next client in the chain is holding.
This is incompatible change, as can be seen by number of tests
cases that needed to be fixed, but is consistent with respect to
behavior of the different scenarios in which the circular wait
might happen.
An unnecessarily restrictive lock were taken on sub-SELECTs during DELETE.
During parsing, a global structure is reused for sub-SELECTs and the attribute
keeping track of lock options were not reset properly.
This patch introduces a new attribute to keep track on the syntactical lock
option elements found in a sub-SELECT and then sets the lock options accordingly.
Now the sub-SELECTs will try to acquire a READ lock if possible
instead of a WRITE lock as inherited from the outer DELETE statement.
SHOW CREATE TABLE fails
Underlying table names, that merge engine fails to open were not
reported.
With this fix CHECK TABLE issued against merge table reports all
underlying table names that it fails to open. Other statements
are unaffected, that is underlying table names are not included
into error message.
This fix doesn't solve SHOW CREATE TABLE issue.
The problem was that some facilities (like CONVERT_TZ() function or
server HELP statement) may require implicit access to some tables in
'mysql' database. This access was done by ordinary means of adding
such tables to the list of tables the query is going to open.
However, if we issued LOCK TABLES before that, we would get "table
was not locked" error trying to open such implicit tables.
The solution is to treat certain tables as MySQL system tables, like
we already do for mysql.proc. Such tables may be opened for reading
at any moment regardless of any locks in effect. The cost of this is
that system table may be locked for writing only together with other
system tables, it is disallowed to lock system tables for writing and
have any other lock on any other table.
After this patch the following tables are treated as MySQL system
tables:
mysql.help_category
mysql.help_keyword
mysql.help_relation
mysql.help_topic
mysql.proc (it already was)
mysql.time_zone
mysql.time_zone_leap_second
mysql.time_zone_name
mysql.time_zone_transition
mysql.time_zone_transition_type
These tables are now opened with open_system_tables_for_read() and
closed with close_system_tables(), or one table may be opened with
open_system_table_for_update() and closed with close_thread_tables()
(the latter is used for mysql.proc table, which is updated as part of
normal MySQL server operation). These functions may be used when
some tables were opened and locked already.
NOTE: online update of time zone tables is not possible during
replication, because there's no time zone cache flush neither on LOCK
TABLES, nor on FLUSH TABLES, so the master may serve stale time zone
data from cache, while on slave updated data will be loaded from the
time zone tables.
and some SP-related cleanups.
- We don't have separate stage for calculation of list of tables
to be prelocked and doing implicit LOCK/UNLOCK any more.
Instead we calculate this list at open_tables() and do implicit
LOCK in lock_tables() (and UNLOCK in close_thread_tables()).
Also now we support cases when same table (with same alias) is
used several times in the same query in SP.
- Cleaned up execution of SP. Moved all common code which handles
LEX and does preparations before statement execution or complex
expression evaluation to auxilary sp_lex_keeper class. Now
all statements in SP (and corresponding instructions) that
evaluate expression which can contain subquery have their
own LEX.
Collect all tables and SPs refered by a statement, and open all tables
with an implicit LOCK TABLES. Do find things refered by triggers and views,
we open them first (and then repeat this until nothing new is found), before
doing the actual lock tables.
Added more DBUG statements
Ensure that we are comparing end space with BINARY strings
Use 'any_db' instead of '' to mean any database. (For HANDLER command)
Only strip ' ' when comparing CHAR, not other space-like characters (like \t)
bmove_allign -> bmove_align
Added OLAP function ROLLUP
Split mysql_fix_privilege_tables to a script and a .sql data file
Added new (MEMROOT*) functions to avoid calling current_thd() when creating some common objects.
Added table_alias_charset, for easier --lower-case-table-name handling
Better SQL_MODE handling (Setting complex options also sets sub options)
New (faster) assembler string functions for x86
added support for quiet
increased line buffer size
client/mysqltest.c
fixed memory leak
added query logging to result file
added error message logging to result file
added enable_query_log/disable_query_log
mysql-test/mysql-test-run.sh
converted tests to use mysqlmanager
Updated test results