------------------------------------------------------------
revno: 2476.1116.1
committer: davi@mysql.com/endora.local
timestamp: Fri 2007-12-14 10:10:19 -0200
message:
DROP TABLE under LOCK TABLES simultaneous to a FLUSH TABLES
WITH READ LOCK (global read lock) can lead to a deadlock.
The solution is to not wait for the global read lock if the
thread is holding any locked tables.
Related to bugs 23713 and 32395. This issues is being fixed
only on 6.0 because it depends on the fix for bug 25858 --
which was fixed only on 6.0.
------------------------------------------------------------
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.
The additional patch. That 'loadxml.test' failure was actually about our testing system,
not the code.
Firstly we need a new mysqltest command, wich i called 'send_eval'. So the expression
can be evaluated, then started in a parallel thread. We only have separane 'send' and
'eval' commands at the moment.
Then we need to add the waiting code after the 'KILL' to our test, so the thread will be killed
before the test goes further. The present 'reap' command doesn't handle the killed threads
well.
per-file comments:
client/mysqltest.cc
Bug#42520 killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
The 'send_eval' command implemented.
mysql-test/r/loadxml.result
Bug#42520 killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
test result updated.
mysql-test/t/loadxml.test
Bug#42520 killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
test case added.
------------------------------------------------------------
revno: 2476.784.4
revision-id: sp1r-davi@moksha.local-20071008114751-46069
parent: sp1r-davi@moksha.local-20071003002731-48537
committer: davi@moksha.local
timestamp: Mon 2007-10-08 08:47:51 -0300
message:
Bug#27249 table_wild with alias: select t1.* as something
Aliases to table wildcards are silently ignored, but they should
not be allowed as it is non-standard and currently useless. There
is not point in having a alias to a wildcard of column names.
The solution is to rewrite the select_item rule so that aliases
for table wildcards are not accepted.
Contribution by Martin Friebe
------------------------------------------------------------
revno: 2597.4.17
revision-id: sp1r-davi@mysql.com/endora.local-20080328174753-24337
parent: sp1r-anozdrin/alik@quad.opbmk-20080328140038-16479
committer: davi@mysql.com/endora.local
timestamp: Fri 2008-03-28 14:47:53 -0300
message:
Bug#15192 "fatal errors" are caught by handlers in stored procedures
The problem is that fatal errors (e.g.: out of memory) were being
caught by stored procedure exception handlers which could cause
the execution to not be stopped due to a continue handler.
The solution is to not call any exception handler if the error is
fatal and send the fatal error to the client.
------------------------------------------------------------
revno: 2597.37.3
revision-id: sp1r-davi@mysql.com/endora.local-20080328123626-16430
parent: sp1r-anozdrin/alik@quad.opbmk-20080327125300-11290
committer: davi@mysql.com/endora.local
timestamp: Fri 2008-03-28 09:36:26 -0300
message:
Bug#10374 GET_LOCK does not let connection to close on the server side if it's aborted
The problem is that the server doesn't detect aborted connections which
are waiting on a lock or sleeping (user sleep), wasting system resources
for a connection that is already dead.
The solution is to peek at the connection every five seconds to verify if
the connection is not aborted. A aborted connection is detect by polling
the connection socket for available data to be read or end of file and in
case of eof, the wait is aborted and the connection killed.
------------------------------------------------------------
revno: 2572.2.1
revision-id: sp1r-davi@mysql.com/endora.local-20080227225948-16317
parent: sp1r-anozdrin/alik@quad.-20080226165712-10409
committer: davi@mysql.com/endora.local
timestamp: Wed 2008-02-27 19:59:48 -0300
message:
Bug#27525 table not found when using multi-table-deletes with aliases over several databas
Bug#30234 Unexpected behavior using DELETE with AS and USING
The multi-delete statement has a documented limitation that
cross-database multiple-table deletes using aliases are not
supported because it fails to find the tables by alias if it
belongs to a different database. The problem is that when
building the list of tables to delete from, if a database
name is not specified (maybe an alias) it defaults to the
name of the current selected database, making impossible to
to properly resolve tables by alias later. Another problem
is a inconsistency of the multiple table delete syntax that
permits ambiguities in a delete statement (aliases that refer
to multiple different tables or vice-versa).
The first step for a solution and proper implementation of
the cross-databse multiple table delete is to get rid of any
ambiguities in a multiple table statement. Currently, the parser
is accepting multiple table delete statements that have no obvious
meaning, such as:
DELETE a1 FROM db1.t1 AS a1, db2.t2 AS a1;
DELETE a1 AS a1 FROM db1.t1 AS a1, db2.t2 AS a1;
The solution is to resolve the left part of a delete statement
using the right part, if the a table on right has an alias,
it must be referenced in the left using the given alias. Also,
each table on the left side must match unambiguously only one
table in the right side.
------------------------------------------------------------
revno: 2630.39.3
revision-id: davi.arnaut@sun.com-20081210215359-i876m4zgc2d6rzs3
parent: kostja@sun.com-20081208222938-9es7wl61moli71ht
committer: Davi Arnaut <Davi.Arnaut@Sun.COM>
branch nick: 36649-6.0
timestamp: Wed 2008-12-10 19:53:59 -0200
message:
Bug#36649: Condition area is not properly cleaned up after stored routine invocation
The problem is that the diagnostics area of a trigger is not
isolated from the area of the statement that caused the trigger
invocation. In MySQL terms, it means that warnings generated
during the execution of the trigger are not removed from the
"warning area" at the end of the execution.
Before this fix, the rules for MySQL message list life cycle (see
manual entry for SHOW WARNINGS) did not apply to statements
inside stored programs:
- The manual says that the list of messages is cleared by a
statement that uses a table (any table). However, such
statement, if run inside a stored program did not clear the
message list.
- The manual says that the list is cleared by a statement that
generates a new error or a warning, but this was not the case
with stored program statements either and is changed to be the
case as well.
In other words, after this fix, a statement has the same effect
on the message list regardless of whether it's executed inside a
stored program/sub-statement or not.
This introduces an incompatible change:
- before this fix, a, e.g. statement inside a trigger could
never clear the global warning list
- after this fix, a trigger that generates a warning or uses a
table, clears the global warning list
- however, when we leave a trigger or a function, the caller's
warning information is restored (see more on this below).
This change is not backward compatible as it is intended to make
MySQL behavior similar to the SQL standard behavior:
A stored function or trigger will get its own "warning area" (or,
in standard terminology, diagnostics area). At the beginning of
the stored function or trigger, all messages from the caller area
will be copied to the area of the trigger. During execution, the
message list will be cleared according to the MySQL rules
described on the manual (SHOW WARNINGS entry). At the end of the
function/trigger, the "warning area" will be destroyed along with
all warnings it contains, except that if the last statement of
the function/trigger generated messages, these are copied into
the "warning area" of the caller.
Consequently, statements that use a table or generate a warning
*will* clear warnings inside the trigger, but that will have no
effect to the warning list of the calling (outer) statement.
Added this option, named as "--dump-slave". The purpose of this option is to be
able to produce a dump from a slave used for making backups of the master. Originally,
dumping from the main master was fine, but as more data accumulated, the dump process
would take over 30 minutes, locking up the master database hence website for 30 minutes.
A slave dedicated to producing backups was the answer, but I needed a dump that could be
used to restore a slave instantly and in order to do that, it has to have three things
contained in the dump:
1. "STOP SLAVE;" at the beginning
2. "CHANGE MASTER TO ...<the master - info from 'show slave status'>"
3. "START SLAVE;" at the end
These options in this changeset contain this.
--stop-slave adds "STOP SLAVE" to the beginning of the dump and "STOP SLAVE"
to the end of the dump.
--include-host gives the user the option to have the host explicitely added
to the "CHANGE MASTER TO ..." line.
--dump-slave adds the "CHANGE MASTER ..." to the dump representing not the slave's
master binlog info, but the slave's master's info from "SHOW SLAVE STATUS"
Feature from Eric Bergen, CLA signed 2007-06-27.
Adds new mysql client option "--auto-vertical-output", which causes
the client to test whether a result table is too wide for the current
window (where available) and emit vertical results in that case.
Otherwise, it sends normal tabular results.
deadlocks
Backport of revno: 2617.68.35
The problem was that if one connection is running a multi-statement
transaction which involves a single partitioned table, and another
connection attempts to alter the table to drop a non-existing partition,
(which of course will fail), the first connection still gets
ER_LOCK_DEADLOCK and cannot proceed anymore.
This bug is no longer reproducable. This has also been tested with the
patch for Bug#46654 "False deadlock on concurrent DML/DDL with partitions,
inconsistent behavior" which concerned a similar problem but where the
ALTER TABLE is semantically correct.
Test case added in partition_sync.test.
Bug#42662: maketime() and signedness
Item_time_typecast::val_int() dropped sign from
MYSQL_TIME gotten using from get_time().
Propagates sign now.
Backported to 5.5.0 (6.0-codebase revid: 1810.3897.1)
When less than six places are given for microseconds, we zerofill from
the right (leftmost place is always 1/10s). We only did this when all
announced date/time fields were given; now we also format fractional
seconds when more significant fields are left out.
Bug#41756 "Strange error messages about locks from InnoDB".
In JT_EQ_REF (join_read_key()) access method,
don't try to unlock rows in the handler, unless certain that
a) they were locked
b) they are not used.
Unlocking of rows is done by the logic of the nested join loop,
and is unaware of the possible caching that the access method may
have. This could lead to double unlocking, when a row
was unlocked first after reading into the cache, and then
when taken from cache, as well as to unlocking of rows which
were actually used (but taken from cache).
Delegate part of the unlocking logic to the access method,
and in JT_EQ_REF count how many times a record was actually
used in the join. Unlock it only if it's usage count is 0.
Implemented review comments.
Bug#41756 "Strange error messages about locks from InnoDB".
In JT_EQ_REF (join_read_key()) access method,
don't try to unlock rows in the handler, unless certain that
a) they were locked
b) they are not used.
Unlocking of rows is done by the logic of the nested join loop,
and is unaware of the possible caching that the access method may
have. This could lead to double unlocking, when a row
was unlocked first after reading into the cache, and then
when taken from cache, as well as to unlocking of rows which
were actually used (but taken from cache).
Delegate part of the unlocking logic to the access method,
and in JT_EQ_REF count how many times a record was actually
used in the join. Unlock it only if it's usage count is 0.
Implemented review comments.
values return too many records
WHERE clauses with "outer_value_list NOT IN subselect" were
handled incorrectly if the outer value list contained multiple
items where at least one of these could be NULL. The first
outer record with NULL value was handled correctly, but if a
second record with NULL value existed, the optimizer would
choose to reuse the result it got on the last execution of the
subselect. This is incorrect if the outer value list has
multiple items.
The fix is to make Item_in_optimizer::val_int (in
item_cmpfunc.cc) reuse the result of the latest execution
for NULL values only if all values in the outer_value_list
are NULL.
