Skip to content

M
MariaDB
Commit 78c6a8ca authored by Dmitry Lenev's avatar Dmitry Lenev
Browse files
Options

A 5.1-only version of fix for bug #46947 "Embedded SELECT 

without FOR UPDATE is causing a lock".

SELECT statements with subqueries referencing InnoDB tables
were acquiring shared locks on rows in these tables when they
were executed in REPEATABLE-READ mode and with statement or
mixed mode binary logging turned on.

This was a regression which were introduced when fixing
bug 39843.

The problem was that for tables belonging to subqueries
parser set TL_READ_DEFAULT as a lock type. In cases when
statement/mixed binary logging at open_tables() time this
type of lock was converted to TL_READ_NO_INSERT lock at
open_tables() time and caused InnoDB engine to acquire
shared locks on reads from these tables. Although in some
cases such behavior was correct (e.g. for subqueries in
DELETE) in case of SELECT it has caused unnecessary locking.

This patch implements minimal version of the fix for the
specific problem described in the bug-report which supposed
to be not too risky for pushing into 5.1 tree.
The 5.5 tree already contains a more appropriate solution
which also addresses other related issues like bug 53921
"Wrong locks for SELECTs used stored functions may lead
to broken SBR".

This patch tries to solve the problem by ensuring that
TL_READ_DEFAULT lock which is set in the parser for
tables participating in subqueries at open_tables()
time is interpreted as TL_READ_NO_INSERT or TL_READ.
TL_READ is used only if we know that this is a SELECT
and that this particular table is not used by a stored
function.

Test coverage is added for both InnoDB and MyISAM.

This patch introduces an "incompatible" change in locking
scheme for subqueries used in SELECT ... FOR UPDATE and
SELECT .. IN SHARE MODE.

In 4.1 (as well as in 5.0 and 5.1 before fix for bug 39843)
the server would use a snapshot InnoDB read for subqueries
in SELECT FOR UPDATE and SELECT .. IN SHARE MODE statements,
regardless of whether the binary log is on or off.

If the user required a different type of read (i.e. locking
read), he/she could request so explicitly by providing FOR
UPDATE/IN SHARE MODE clause for each individual subquery.

The patch for bug 39843 broke this behaviour (which was not
documented or tested), and started to use locking reads for
all subqueries in SELECT ... FOR UPDATE/IN SHARE MODE.
This patch restores 4.1 behaviour.

This patch should be mostly null-merged into 5.5 tree.
parent fa3570f9
Show whitespace changes
Inline Side-by-side
Showing with 3245 additions and 17 deletions
mysql-test/include/check_concurrent_insert.inc 0 → 100644
View file @ 78c6a8ca
#
# SUMMARY
# Check if statement reading table '$table' allows concurrent
# inserts in it.
#
# PARAMETERS
# $table Table in which concurrent inserts should be allowed.
# $con_aux1 Name of the first auxiliary connection to be used by this
# script.
# $con_aux2 Name of the second auxiliary connection to be used by this
# script.
# $statement Statement to be checked.
# $restore_table Table which might be modified by statement to be checked
# and thus needs backing up before its execution and
# restoring after it (can be empty).
#
# EXAMPLE
# lock_sync.test
#
--disable_result_log
--disable_query_log
# Reset DEBUG_SYNC facility for safety.
set debug_sync= "RESET";
if (`SELECT '$restore_table' <> ''`)
{
--eval create temporary table t_backup select * from $restore_table;
}
connection $con_aux1;
set debug_sync='after_lock_tables_takes_lock SIGNAL parked WAIT_FOR go';
--send_eval $statement;
connection $con_aux2;
set debug_sync='now WAIT_FOR parked';
--send_eval insert into $table (i) values (0);
--enable_result_log
--enable_query_log
connection default;
# Wait until concurrent insert is successfully executed while
# statement being checked has its tables locked.
# We use wait_condition.inc instead of simply reaping
# concurrent insert here in order to avoid deadlocks if test
# fails and to time out gracefully instead.
let $wait_condition=
select count(*) = 0 from information_schema.processlist
where info = "insert into $table (i) values (0)";
--source include/wait_condition.inc
--disable_result_log
--disable_query_log
if ($success)
{
# Apparently concurrent insert was successfully executed.
# To be safe against wait_condition.inc succeeding due to
# races let us first reap concurrent insert to ensure that
# it has really been successfully executed.
connection $con_aux2;
--reap
connection default;
set debug_sync= 'now SIGNAL go';
connection $con_aux1;
--reap
connection default;
--echo Success: '$statement' allows concurrent inserts into '$table'.
}
if (!$success)
{
# Waiting has timed out. Apparently concurrent insert was blocked.
# So to be able to continue we need to end our statement first.
set debug_sync= 'now SIGNAL go';
connection $con_aux1;
--reap
connection $con_aux2;
--reap
connection default;
--echo Error: '$statement' doesn't allow concurrent inserts into '$table'!
}
--eval delete from $table where i = 0;
if (`SELECT '$restore_table' <> ''`)
{
--eval truncate table $restore_table;
--eval insert into $restore_table select * from t_backup;
drop temporary table t_backup;
}
# Clean-up. Reset DEBUG_SYNC facility after use.
set debug_sync= "RESET";
--enable_result_log
--enable_query_log
mysql-test/include/check_no_concurrent_insert.inc 0 → 100644
View file @ 78c6a8ca
#
# SUMMARY
# Check that statement reading table '$table' doesn't allow concurrent
# inserts in it.
#
# PARAMETERS
# $table Table in which concurrent inserts should be disallowed.
# $con_aux1 Name of the first auxiliary connection to be used by this
# script.
# $con_aux2 Name of the second auxiliary connection to be used by this
# script.
# $statement Statement to be checked.
# $restore_table Table which might be modified by statement to be checked
# and thus needs backing up before its execution and
# restoring after it (can be empty).
#
# EXAMPLE
# lock_sync.test
#
--disable_result_log
--disable_query_log
# Reset DEBUG_SYNC facility for safety.
set debug_sync= "RESET";
if (`SELECT '$restore_table' <> ''`)
{
--eval create temporary table t_backup select * from $restore_table;
}
connection $con_aux1;
set debug_sync='after_lock_tables_takes_lock SIGNAL parked WAIT_FOR go';
--send_eval $statement;
connection $con_aux2;
set debug_sync='now WAIT_FOR parked';
--send_eval insert into $table (i) values (0);
--enable_result_log
--enable_query_log
connection default;
# Wait until concurrent insert is successfully blocked because
# of our statement.
let $wait_condition=
select count(*) = 1 from information_schema.processlist
where state = "Locked" and info = "insert into $table (i) values (0)";
--source include/wait_condition.inc
--disable_result_log
--disable_query_log
set debug_sync= 'now SIGNAL go';
connection $con_aux1;
--reap
connection $con_aux2;
--reap
connection default;
if ($success)
{
--echo Success: '$statement' doesn't allow concurrent inserts into '$table'.
}
if (!$success)
{
--echo Error: '$statement' allows concurrent inserts into '$table'!
}
--eval delete from $table where i = 0;
if (`SELECT '$restore_table' <> ''`)
{
--eval truncate table $restore_table;
--eval insert into $restore_table select * from t_backup;
drop temporary table t_backup;
}
# Clean-up. Reset DEBUG_SYNC facility after use.
set debug_sync= "RESET";
--enable_result_log
--enable_query_log
mysql-test/include/check_no_row_lock.inc 0 → 100644
View file @ 78c6a8ca
#
# SUMMARY
# Check if statement affecting or reading table '$table' doesn't
# take any kind of locks on its rows.
#
# PARAMETERS
# $table Table for which presence of row locks should be checked.
# $con_aux Name of auxiliary connection to be used by this script.
# $statement Statement to be checked.
#
# EXAMPLE
# innodb_mysql_lock2.test
#
--disable_result_log
--disable_query_log
connection default;
begin;
--eval select * from $table for update;
connection $con_aux;
begin;
--send_eval $statement;
--enable_result_log
--enable_query_log
connection default;
# Wait until statement is successfully executed while
# all rows in table are X-locked. This means that it
# does not acquire any row locks.
# We use wait_condition.inc instead of simply reaping
# statement here in order to avoid deadlocks if test
# fails and to time out gracefully instead.
let $wait_condition=
select count(*) = 0 from information_schema.processlist
where info = "$statement";
--source include/wait_condition.inc
--disable_result_log
--disable_query_log
if ($success)
{
# Apparently statement was successfully executed and thus it
# has not required any row locks.
# To be safe against wait_condition.inc succeeding due to
# races let us first reap the statement being checked to
# ensure that it has been successfully executed.
connection $con_aux;
--reap
rollback;
connection default;
rollback;
--echo Success: '$statement' doesn't take row locks on '$table'.
}
if (!$success)
{
# Waiting has timed out. Apparently statement was blocked on
# some row lock. So to be able to continue we need to unlock
# rows first.
rollback;
connection $con_aux;
--reap
rollback;
connection default;
--echo Error: '$statement' takes some row locks on '$table'!
}
--enable_result_log
--enable_query_log
mysql-test/include/check_shared_row_lock.inc 0 → 100644
View file @ 78c6a8ca
#
# SUMMARY
# Check if statement reading table '$table' takes shared locks
# on some of its rows.
#
# PARAMETERS
# $table Table for which presence of row locks should be checked.
# $con_aux Name of auxiliary connection to be used by this script.
# $statement Statement to be checked.
# $wait_statement Sub-statement which is supposed to acquire locks (should
# be the same as $statement for ordinary statements).
#
# EXAMPLE
# innodb_mysql_lock2.test
#
--disable_result_log
--disable_query_log
connection default;
begin;
--eval select * from $table for update;
connection $con_aux;
begin;
--send_eval $statement;
--enable_result_log
--enable_query_log
connection default;
# Wait until statement is successfully blocked because
# all rows in table are X-locked. This means that at
# least it acquires S-locks on some of rows.
let $wait_condition=
select count(*) = 1 from information_schema.processlist
where state in ("Sending data","statistics", "preparing") and
info = "$wait_statement";
--source include/wait_condition.inc
--disable_result_log
--disable_query_log
rollback;
connection $con_aux;
--reap
rollback;
connection default;
--enable_result_log
--enable_query_log
if ($success)
{
--echo Success: '$statement' takes shared row locks on '$table'.
}
if (!$success)
{
--echo Error: '$statement' hasn't taken shared row locks on '$table'!
}
mysql-test/r/bug39022.result
View file @ 78c6a8ca
...@@ -12,7 +12,7 @@ INSERT INTO t2 VALUES (0),(1),(2),(3),(4),(5),(6),(7),(8),(9),(10), ...@@ -12,7 +12,7 @@ INSERT INTO t2 VALUES (0),(1),(2),(3),(4),(5),(6),(7),(8),(9),(10),
START TRANSACTION; START TRANSACTION;
# in thread2 # in thread2
REPLACE INTO t2 VALUES (-17); REPLACE INTO t2 VALUES (-17);
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE;
d d
# in thread1 # in thread1
REPLACE INTO t1(a,b) VALUES (67,20); REPLACE INTO t1(a,b) VALUES (67,20);
...@@ -21,10 +21,10 @@ COMMIT; ...@@ -21,10 +21,10 @@ COMMIT;
START TRANSACTION; START TRANSACTION;
REPLACE INTO t1(a,b) VALUES (65,-50); REPLACE INTO t1(a,b) VALUES (65,-50);
REPLACE INTO t2 VALUES (-91); REPLACE INTO t2 VALUES (-91);
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE;
# in thread1 # in thread1
# should not crash # should not crash
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE;
ERROR 40001: Deadlock found when trying to get lock; try restarting transaction ERROR 40001: Deadlock found when trying to get lock; try restarting transaction
# in thread2 # in thread2
d d
......
mysql-test/r/innodb_mysql_lock2.result 0 → 100644
View file @ 78c6a8ca
#
# Test how do we handle locking in various cases when
# we read data from InnoDB tables.
#
# In fact by performing this test we check two things:
# 1) That SQL-layer correctly determine type of thr_lock.c
# lock to be acquired/passed to InnoDB engine.
# 2) That InnoDB engine correctly interprets this lock
# type and takes necessary row locks or does not
# take them if they are not necessary.
#
# This test makes sense only in REPEATABLE-READ mode as
# in SERIALIZABLE mode all statements that read data take
# shared lock on them to enforce its semantics.
select @@session.tx_isolation;
@@session.tx_isolation
REPEATABLE-READ
# Prepare playground by creating tables, views,
# routines and triggers used in tests.
drop table if exists t0, t1, t2, t3, t4, t5, te;
drop view if exists v1, v2;
drop procedure if exists p1;
drop procedure if exists p2;
drop function if exists f1;
drop function if exists f2;
drop function if exists f3;
drop function if exists f4;
drop function if exists f5;
drop function if exists f6;
drop function if exists f7;
drop function if exists f8;
drop function if exists f9;
drop function if exists f10;
drop function if exists f11;
drop function if exists f12;
drop function if exists f13;
drop function if exists f14;
drop function if exists f15;
create table t1 (i int primary key) engine=innodb;
insert into t1 values (1), (2), (3), (4), (5);
create table t2 (j int primary key) engine=innodb;
insert into t2 values (1), (2), (3), (4), (5);
create table t3 (k int primary key) engine=innodb;
insert into t3 values (1), (2), (3);
create table t4 (l int primary key) engine=innodb;
insert into t4 values (1);
create table t5 (l int primary key) engine=innodb;
insert into t5 values (1);
create table te(e int primary key);
insert into te values (1);
create view v1 as select i from t1;
create view v2 as select j from t2 where j in (select i from t1);
create procedure p1(k int) insert into t2 values (k);
create function f1() returns int
begin
declare j int;
select i from t1 where i = 1 into j;
return j;
end|
create function f2() returns int
begin
declare k int;
select i from t1 where i = 1 into k;
insert into t2 values (k + 5);
return 0;
end|
create function f3() returns int
begin
return (select i from t1 where i = 3);
end|
create function f4() returns int
begin
if (select i from t1 where i = 3) then
return 1;
else
return 0;
end if;
end|
create function f5() returns int
begin
insert into t2 values ((select i from t1 where i = 1) + 5);
return 0;
end|
create function f6() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
return k;
end|
create function f7() returns int
begin
declare k int;
select j from v2 where j = 1 into k;
return k;
end|
create function f8() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
insert into t2 values (k+5);
return k;
end|
create function f9() returns int
begin
update v2 set j=j+10 where j=1;
return 1;
end|
create function f10() returns int
begin
return f1();
end|
create function f11() returns int
begin
declare k int;
set k= f1();
insert into t2 values (k+5);
return k;
end|
create function f12(p int) returns int
begin
insert into t2 values (p);
return p;
end|
create function f13(p int) returns int
begin
return p;
end|
create procedure p2(inout p int)
begin
select i from t1 where i = 1 into p;
end|
create function f14() returns int
begin
declare k int;
call p2(k);
insert into t2 values (k+5);
return k;
end|
create function f15() returns int
begin
declare k int;
call p2(k);
return k;
end|
create trigger t4_bi before insert on t4 for each row
begin
declare k int;
select i from t1 where i=1 into k;
set new.l= k+1;
end|
create trigger t4_bu before update on t4 for each row
begin
if (select i from t1 where i=1) then
set new.l= 2;
end if;
end|
# Trigger below uses insertion of duplicate key in 'te'
# table as a way to abort delete operation.
create trigger t4_bd before delete on t4 for each row
begin
if !(select i from v1 where i=1) then
insert into te values (1);
end if;
end|
create trigger t5_bi before insert on t5 for each row
begin
set new.l= f1()+1;
end|
create trigger t5_bu before update on t5 for each row
begin
declare j int;
call p2(j);
set new.l= j + 1;
end|
#
# Set common variables to be used by scripts called below.
#
#
# 1. Statements that read tables and do not use subqueries.
#
#
# 1.1 Simple SELECT statement.
#
# No locks are necessary as this statement won't be written
# to the binary log and InnoDB supports snapshots.
Success: 'select * from t1' doesn't take row locks on 't1'.
#
# 1.2 Multi-UPDATE statement.
#
# Has to take shared locks on rows in the table being read as this
# statement will be written to the binary log and therefore should
# be serialized with concurrent statements.
Success: 'update t2, t1 set j= j - 1 where i = j' takes shared row locks on 't1'.
#
# 1.3 Multi-DELETE statement.
#
# The above is true for this statement as well.
Success: 'delete t2 from t1, t2 where i = j' takes shared row locks on 't1'.
#
# 1.4 DESCRIBE statement.
#
# This statement does not really read data from the
# target table and thus does not take any lock on it.
# We check this for completeness of coverage.
Success: 'describe t1' doesn't take row locks on 't1'.
#
# 1.5 SHOW statements.
#
# The above is true for SHOW statements as well.
Success: 'show create table t1' doesn't take row locks on 't1'.
Success: 'show keys from t1' doesn't take row locks on 't1'.
#
# 2. Statements which read tables through subqueries.
#
#
# 2.1 CALL with a subquery.
#
# A strong lock is not necessary as this statement is not
# written to the binary log as a whole (it is written
# statement-by-statement) and thanks to MVCC we can always get
# versions of rows prior to the update that has locked them.
# But in practice InnoDB does locking reads for all statements
# other than SELECT (unless it is a READ-COMITTED mode or
# innodb_locks_unsafe_for_binlog is ON).
Success: 'call p1((select i + 5 from t1 where i = 1))' takes shared row locks on 't1'.
#
# 2.2 CREATE TABLE with a subquery.
#
# Has to take shared locks on rows in the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent statements.
Success: 'create table t0 engine=innodb select * from t1' takes shared row locks on 't1'.
drop table t0;
Success: 'create table t0 engine=innodb select j from t2 where j in (select i from t1)' takes shared row locks on 't1'.
drop table t0;
#
# 2.3 DELETE with a subquery.
#
# The above is true for this statement as well.
Success: 'delete from t2 where j in (select i from t1)' takes shared row locks on 't1'.
#
# 2.4 MULTI-DELETE with a subquery.
#
# Same is true for this statement as well.
Success: 'delete t2 from t3, t2 where k = j and j in (select i from t1)' takes shared row locks on 't1'.
#
# 2.5 DO with a subquery.
#
# In theory should not take row locks as it is not logged.
# In practice InnoDB takes shared row locks.
Success: 'do (select i from t1 where i = 1)' takes shared row locks on 't1'.
#
# 2.6 INSERT with a subquery.
#
# Has to take shared locks on rows in the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent statements.
Success: 'insert into t2 select i+5 from t1' takes shared row locks on 't1'.
Success: 'insert into t2 values ((select i+5 from t1 where i = 4))' takes shared row locks on 't1'.
#
# 2.7 LOAD DATA with a subquery.
#
# The above is true for this statement as well.
Success: 'load data infile '../../std_data/rpl_loaddata.dat' into table t2 (@a, @b) set j= @b + (select i from t1 where i = 1)' takes shared row locks on 't1'.
#
# 2.8 REPLACE with a subquery.
#
# Same is true for this statement as well.
Success: 'replace into t2 select i+5 from t1' takes shared row locks on 't1'.
Success: 'replace into t2 values ((select i+5 from t1 where i = 4))' takes shared row locks on 't1'.
#
# 2.9 SELECT with a subquery.
#
# Locks are not necessary as this statement is not written
# to the binary log and thanks to MVCC we can always get
# versions of rows prior to the update that has locked them.
#
# Also serves as a test case for bug #46947 "Embedded SELECT
# without FOR UPDATE is causing a lock".
Success: 'select * from t2 where j in (select i from t1)' doesn't take row locks on 't1'.
#
# 2.10 SET with a subquery.
#
# In theory should not require locking as it is not written
# to the binary log. In practice InnoDB acquires shared row
# locks.
Success: 'set @a:= (select i from t1 where i = 1)' takes shared row locks on 't1'.
#
# 2.11 SHOW with a subquery.
#
# Similarly to the previous case, in theory should not require locking
# as it is not written to the binary log. In practice InnoDB
# acquires shared row locks.
Success: 'show tables from test where Tables_in_test = 't2' and (select i from t1 where i = 1)' takes shared row locks on 't1'.
Success: 'show columns from t2 where (select i from t1 where i = 1)' takes shared row locks on 't1'.
#
# 2.12 UPDATE with a subquery.
#
# Has to take shared locks on rows in the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent statements.
Success: 'update t2 set j= j-10 where j in (select i from t1)' takes shared row locks on 't1'.
#
# 2.13 MULTI-UPDATE with a subquery.
#
# Same is true for this statement as well.
Success: 'update t2, t3 set j= j -10 where j=k and j in (select i from t1)' takes shared row locks on 't1'.
#
# 3. Statements which read tables through a view.
#
#
# 3.1 SELECT statement which uses some table through a view.
#
# Since this statement is not written to the binary log
# and old version of rows are accessible thanks to MVCC,
# no locking is necessary.
Success: 'select * from v1' doesn't take row locks on 't1'.
Success: 'select * from v2' doesn't take row locks on 't1'.
Success: 'select * from t2 where j in (select i from v1)' doesn't take row locks on 't1'.
Success: 'select * from t3 where k in (select j from v2)' doesn't take row locks on 't1'.
#
# 3.2 Statements which modify a table and use views.
#
# Since such statements are going to be written to the binary
# log they need to be serialized against concurrent statements
# and therefore should take shared row locks on data read.
Success: 'update t2 set j= j-10 where j in (select i from v1)' takes shared row locks on 't1'.
Success: 'update t3 set k= k-10 where k in (select j from v2)' takes shared row locks on 't1'.
Success: 'update t2, v1 set j= j-10 where j = i' takes shared row locks on 't1'.
Success: 'update v2 set j= j-10 where j = 3' takes shared row locks on 't1'.
#
# 4. Statements which read tables through stored functions.
#
#
# 4.1 SELECT/SET with a stored function which does not
# modify data and uses SELECT in its turn.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f1()' doesn't take row locks on 't1'.
Success: 'set @a:= f1()' doesn't take row locks on 't1'.
#
# 4.2 INSERT (or other statement which modifies data) with
# a stored function which does not modify data and uses
# SELECT.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting the data
# it uses. Therefore it should take row locks on the data
# it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'insert into t2 values (f1() + 5)' doesn't take row locks on 't1'.
#
# 4.3 SELECT/SET with a stored function which
# reads and modifies data.
#
# Since a call to such function is written to the binary log,
# it should be serialized with concurrent statements affecting
# the data it uses. Hence, row locks on the data read
# should be taken.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'select f2()' doesn't take row locks on 't1'.
Success: 'set @a:= f2()' doesn't take row locks on 't1'.
#
# 4.4. SELECT/SET with a stored function which does not
# modify data and reads a table through subselect
# in a control construct.
#
# Again, in theory a call to this function won't get to the
# binary log and thus no locking is needed. But in practice
# we don't detect this fact early enough (get_lock_type_for_table())
# to avoid taking row locks.
Success: 'select f3()' takes shared row locks on 't1'.
Success: 'set @a:= f3()' takes shared row locks on 't1'.
Success: 'select f4()' takes shared row locks on 't1'.
Success: 'set @a:= f4()' takes shared row locks on 't1'.
#
# 4.5. INSERT (or other statement which modifies data) with
# a stored function which does not modify data and reads
# the table through a subselect in one of its control
# constructs.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting data it
# uses. Therefore it should take row locks on the data
# it reads.
Success: 'insert into t2 values (f3() + 5)' takes shared row locks on 't1'.
Success: 'insert into t2 values (f4() + 6)' takes shared row locks on 't1'.
#
# 4.6 SELECT/SET which uses a stored function with
# DML which reads a table via a subquery.
#
# Since call to such function is written to the binary log
# it should be serialized with concurrent statements.
# Hence reads should take row locks.
Success: 'select f5()' takes shared row locks on 't1'.
Success: 'set @a:= f5()' takes shared row locks on 't1'.
#
# 4.7 SELECT/SET which uses a stored function which
# doesn't modify data and reads tables through
# a view.
#
# Once again, in theory, calls to such functions won't
# get into the binary log and thus don't need row
# locks. In practice this fact is discovered
# too late to have any effect.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken
# in case of simple SELECT.
Success: 'select f6()' doesn't take row locks on 't1'.
Success: 'set @a:= f6()' doesn't take row locks on 't1'.
Success: 'select f7()' takes shared row locks on 't1'.
Success: 'set @a:= f7()' takes shared row locks on 't1'.
#
# 4.8 INSERT which uses stored function which
# doesn't modify data and reads a table
# through a view.
#
# Since such statement is written to the binary log and
# should be serialized with concurrent statements affecting
# the data it uses. Therefore it should take row locks on
# the rows it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken
# in case of simple SELECT.
Success: 'insert into t3 values (f6() + 5)' doesn't take row locks on 't1'.
Success: 'insert into t3 values (f7() + 5)' takes shared row locks on 't1'.
#
# 4.9 SELECT which uses a stored function which
# modifies data and reads tables through a view.
#
# Since a call to such function is written to the binary log
# it should be serialized with concurrent statements.
# Hence, reads should take row locks.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken
# in case of simple SELECT.
Success: 'select f8()' doesn't take row locks on 't1'.
Success: 'select f9()' takes shared row locks on 't1'.
#
# 4.10 SELECT which uses stored function which doesn't modify
# data and reads a table indirectly, by calling another
# function.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f10()' doesn't take row locks on 't1'.
#
# 4.11 INSERT which uses a stored function which doesn't modify
# data and reads a table indirectly, by calling another
# function.
#
# Since such statement is written to the binary log, it should
# be serialized with concurrent statements affecting the data it
# uses. Therefore it should take row locks on data it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'insert into t2 values (f10() + 5)' doesn't take row locks on 't1'.
#
# 4.12 SELECT which uses a stored function which modifies
# data and reads a table indirectly, by calling another
# function.
#
# Since a call to such function is written to the binary log
# it should be serialized from concurrent statements.
# Hence, reads should take row locks.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'select f11()' doesn't take row locks on 't1'.
#
# 4.13 SELECT that reads a table through a subquery passed
# as a parameter to a stored function which modifies
# data.
#
# Even though a call to this function is written to the
# binary log, values of its parameters are written as literals.
# So there is no need to acquire row locks on rows used in
# the subquery.
# But due to the fact that in 5.1 for prelocked statements
# THD::in_lock_tables is set to TRUE we acquire strong locks
# (see also bug#44613 "SELECT statement inside FUNCTION takes
# a shared lock" [sic!!!]).
Success: 'select f12((select i+10 from t1 where i=1))' takes shared row locks on 't1'.
#
# 4.14 INSERT that reads a table via a subquery passed
# as a parameter to a stored function which doesn't
# modify data.
#
# Since this statement is written to the binary log it should
# be serialized with concurrent statements affecting the data it
# uses. Therefore it should take row locks on the data it reads.
Success: 'insert into t2 values (f13((select i+10 from t1 where i=1)))' takes shared row locks on 't1'.
#
# 5. Statements that read tables through stored procedures.
#
#
# 5.1 CALL statement which reads a table via SELECT.
#
# Since neither this statement nor its components are
# written to the binary log, there is no need to take
# row locks on the data it reads.
Success: 'call p2(@a)' doesn't take row locks on 't1'.
#
# 5.2 Function that modifies data and uses CALL,
# which reads a table through SELECT.
#
# Since a call to such function is written to the binary
# log, it should be serialized with concurrent statements.
# Hence, in this case reads should take row locks on data.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'select f14()' doesn't take row locks on 't1'.
#
# 5.3 SELECT that calls a function that doesn't modify data and
# uses a CALL statement that reads a table via SELECT.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f15()' doesn't take row locks on 't1'.
#
# 5.4 INSERT which calls function which doesn't modify data and
# uses CALL statement which reads table through SELECT.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting data it
# uses. Therefore it should take row locks on data it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'insert into t2 values (f15()+5)' doesn't take row locks on 't1'.
#
# 6. Statements that use triggers.
#
#
# 6.1 Statement invoking a trigger that reads table via SELECT.
#
# Since this statement is written to the binary log it should
# be serialized with concurrent statements affecting the data
# it uses. Therefore, it should take row locks on the data
# it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'insert into t4 values (2)' doesn't take row locks on 't1'.
#
# 6.2 Statement invoking a trigger that reads table through
# a subquery in a control construct.
#
# The above is true for this statement as well.
Success: 'update t4 set l= 2 where l = 1' takes shared row locks on 't1'.
#
# 6.3 Statement invoking a trigger that reads a table through
# a view.
#
# And for this statement.
Success: 'delete from t4 where l = 1' takes shared row locks on 't1'.
#
# 6.4 Statement invoking a trigger that reads a table through
# a stored function.
#
# And for this statement.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'insert into t5 values (2)' doesn't take row locks on 't1'.
#
# 6.5 Statement invoking a trigger that reads a table through
# stored procedure.
#
# And for this statement.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" no lock is taken.
Success: 'update t5 set l= 2 where l = 1' doesn't take row locks on 't1'.
# Clean-up.
drop function f1;
drop function f2;
drop function f3;
drop function f4;
drop function f5;
drop function f6;
drop function f7;
drop function f8;
drop function f9;
drop function f10;
drop function f11;
drop function f12;
drop function f13;
drop function f14;
drop function f15;
drop view v1, v2;
drop procedure p1;
drop procedure p2;
drop table t1, t2, t3, t4, t5, te;
mysql-test/r/lock_sync.result 0 → 100644
View file @ 78c6a8ca
#
# Test how we handle locking in various cases when
# we read data from MyISAM tables.
#
# In this test we mostly check that the SQL-layer correctly
# determines the type of thr_lock.c lock for a table being
# read.
# I.e. that it disallows concurrent inserts when the statement
# is going to be written to the binary log and therefore
# should be serialized, and allows concurrent inserts when
# such serialization is not necessary (e.g. when
# the statement is not written to binary log).
#
# Force concurrent inserts to be performed even if the table
# has gaps. This allows to simplify clean up in scripts
# used below (instead of backing up table being inserted
# into and then restoring it from backup at the end of the
# script we can simply delete rows which were inserted).
set @old_concurrent_insert= @@global.concurrent_insert;
set @@global.concurrent_insert= 2;
select @@global.concurrent_insert;
@@global.concurrent_insert
2
# Prepare playground by creating tables, views,
# routines and triggers used in tests.
drop table if exists t0, t1, t2, t3, t4, t5, te;
drop view if exists v1, v2;
drop procedure if exists p1;
drop procedure if exists p2;
drop function if exists f1;
drop function if exists f2;
drop function if exists f3;
drop function if exists f4;
drop function if exists f5;
drop function if exists f6;
drop function if exists f7;
drop function if exists f8;
drop function if exists f9;
drop function if exists f10;
drop function if exists f11;
drop function if exists f12;
drop function if exists f13;
drop function if exists f14;
drop function if exists f15;
create table t1 (i int primary key);
insert into t1 values (1), (2), (3), (4), (5);
create table t2 (j int primary key);
insert into t2 values (1), (2), (3), (4), (5);
create table t3 (k int primary key);
insert into t3 values (1), (2), (3);
create table t4 (l int primary key);
insert into t4 values (1);
create table t5 (l int primary key);
insert into t5 values (1);
create table te(e int primary key);
insert into te values (1);
create view v1 as select i from t1;
create view v2 as select j from t2 where j in (select i from t1);
create procedure p1(k int) insert into t2 values (k);
create function f1() returns int
begin
declare j int;
select i from t1 where i = 1 into j;
return j;
end|
create function f2() returns int
begin
declare k int;
select i from t1 where i = 1 into k;
insert into t2 values (k + 5);
return 0;
end|
create function f3() returns int
begin
return (select i from t1 where i = 3);
end|
create function f4() returns int
begin
if (select i from t1 where i = 3) then
return 1;
else
return 0;
end if;
end|
create function f5() returns int
begin
insert into t2 values ((select i from t1 where i = 1) + 5);
return 0;
end|
create function f6() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
return k;
end|
create function f7() returns int
begin
declare k int;
select j from v2 where j = 1 into k;
return k;
end|
create function f8() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
insert into t2 values (k+5);
return k;
end|
create function f9() returns int
begin
update v2 set j=j+10 where j=1;
return 1;
end|
create function f10() returns int
begin
return f1();
end|
create function f11() returns int
begin
declare k int;
set k= f1();
insert into t2 values (k+5);
return k;
end|
create function f12(p int) returns int
begin
insert into t2 values (p);
return p;
end|
create function f13(p int) returns int
begin
return p;
end|
create procedure p2(inout p int)
begin
select i from t1 where i = 1 into p;
end|
create function f14() returns int
begin
declare k int;
call p2(k);
insert into t2 values (k+5);
return k;
end|
create function f15() returns int
begin
declare k int;
call p2(k);
return k;
end|
create trigger t4_bi before insert on t4 for each row
begin
declare k int;
select i from t1 where i=1 into k;
set new.l= k+1;
end|
create trigger t4_bu before update on t4 for each row
begin
if (select i from t1 where i=1) then
set new.l= 2;
end if;
end|
# Trigger below uses insertion of duplicate key in 'te'
# table as a way to abort delete operation.
create trigger t4_bd before delete on t4 for each row
begin
if !(select i from v1 where i=1) then
insert into te values (1);
end if;
end|
create trigger t5_bi before insert on t5 for each row
begin
set new.l= f1()+1;
end|
create trigger t5_bu before update on t5 for each row
begin
declare j int;
call p2(j);
set new.l= j + 1;
end|
#
# Set common variables to be used by the scripts
# called below.
#
# Switch to connection 'con1'.
# Cache all functions used in the tests below so statements
# calling them won't need to open and lock mysql.proc table
# and we can assume that each statement locks its tables
# once during its execution.
show create procedure p1;
show create procedure p2;
show create function f1;
show create function f2;
show create function f3;
show create function f4;
show create function f5;
show create function f6;
show create function f7;
show create function f8;
show create function f9;
show create function f10;
show create function f11;
show create function f12;
show create function f13;
show create function f14;
show create function f15;
# Switch back to connection 'default'.
#
# 1. Statements that read tables and do not use subqueries.
#
#
# 1.1 Simple SELECT statement.
#
# No locks are necessary as this statement won't be written
# to the binary log and thanks to how MyISAM works SELECT
# will see version of the table prior to concurrent insert.
Success: 'select * from t1' allows concurrent inserts into 't1'.
#
# 1.2 Multi-UPDATE statement.
#
# Has to take shared locks on rows in the table being read as this
# statement will be written to the binary log and therefore should
# be serialized with concurrent statements.
Success: 'update t2, t1 set j= j - 1 where i = j' doesn't allow concurrent inserts into 't1'.
#
# 1.3 Multi-DELETE statement.
#
# The above is true for this statement as well.
Success: 'delete t2 from t1, t2 where i = j' doesn't allow concurrent inserts into 't1'.
#
# 1.4 DESCRIBE statement.
#
# This statement does not really read data from the
# target table and thus does not take any lock on it.
# We check this for completeness of coverage.
lock table t1 write;
# Switching to connection 'con1'.
# This statement should not be blocked.
describe t1;
# Switching to connection 'default'.
unlock tables;
#
# 1.5 SHOW statements.
#
# The above is true for SHOW statements as well.
lock table t1 write;
# Switching to connection 'con1'.
# These statements should not be blocked.
show keys from t1;
# Switching to connection 'default'.
unlock tables;
#
# 2. Statements which read tables through subqueries.
#
#
# 2.1 CALL with a subquery.
#
# In theory strong lock is not necessary as this statement
# is not written to the binary log as a whole (it is written
# statement-by-statement). But in practice in 5.1 for
# almost everything except SELECT we take strong lock.
Success: 'call p1((select i + 5 from t1 where i = 1))' doesn't allow concurrent inserts into 't1'.
#
# 2.2 CREATE TABLE with a subquery.
#
# Has to take a strong lock on the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent statements.
Success: 'create table t0 select * from t1' doesn't allow concurrent inserts into 't1'.
drop table t0;
Success: 'create table t0 select j from t2 where j in (select i from t1)' doesn't allow concurrent inserts into 't1'.
drop table t0;
#
# 2.3 DELETE with a subquery.
#
# The above is true for this statement as well.
Success: 'delete from t2 where j in (select i from t1)' doesn't allow concurrent inserts into 't1'.
#
# 2.4 MULTI-DELETE with a subquery.
#
# Same is true for this statement as well.
Success: 'delete t2 from t3, t2 where k = j and j in (select i from t1)' doesn't allow concurrent inserts into 't1'.
#
# 2.5 DO with a subquery.
#
# In theory strong lock is not necessary as it is not logged.
# But in practice in 5.1 for almost everything except SELECT
# we take strong lock.
Success: 'do (select i from t1 where i = 1)' doesn't allow concurrent inserts into 't1'.
#
# 2.6 INSERT with a subquery.
#
# Has to take a strong lock on the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent inserts.
Success: 'insert into t2 select i+5 from t1' doesn't allow concurrent inserts into 't1'.
Success: 'insert into t2 values ((select i+5 from t1 where i = 4))' doesn't allow concurrent inserts into 't1'.
#
# 2.7 LOAD DATA with a subquery.
#
# The above is true for this statement as well.
Success: 'load data infile '../../std_data/rpl_loaddata.dat' into table t2 (@a, @b) set j= @b + (select i from t1 where i = 1)' doesn't allow concurrent inserts into 't1'.
#
# 2.8 REPLACE with a subquery.
#
# Same is true for this statement as well.
Success: 'replace into t2 select i+5 from t1' doesn't allow concurrent inserts into 't1'.
Success: 'replace into t2 values ((select i+5 from t1 where i = 4))' doesn't allow concurrent inserts into 't1'.
#
# 2.9 SELECT with a subquery.
#
# Strong locks are not necessary as this statement is not written
# to the binary log and thanks to how MyISAM works this statement
# sees a version of the table prior to the concurrent insert.
Success: 'select * from t2 where j in (select i from t1)' allows concurrent inserts into 't1'.
#
# 2.10 SET with a subquery.
#
# In theory the same is true for this statement as well.
# But in practice in 5.1 we acquire strong lock in this
# case as well.
Success: 'set @a:= (select i from t1 where i = 1)' doesn't allow concurrent inserts into 't1'.
#
# 2.11 SHOW with a subquery.
#
# The same is true for this statement too.
Success: 'show tables from test where Tables_in_test = 't2' and (select i from t1 where i = 1)' doesn't allow concurrent inserts into 't1'.
Success: 'show columns from t2 where (select i from t1 where i = 1)' doesn't allow concurrent inserts into 't1'.
#
# 2.12 UPDATE with a subquery.
#
# Has to take a strong lock on the table being read as
# this statement is written to the binary log and therefore
# should be serialized with concurrent inserts.
Success: 'update t2 set j= j-10 where j in (select i from t1)' doesn't allow concurrent inserts into 't1'.
#
# 2.13 MULTI-UPDATE with a subquery.
#
# Same is true for this statement as well.
Success: 'update t2, t3 set j= j -10 where j=k and j in (select i from t1)' doesn't allow concurrent inserts into 't1'.
#
# 3. Statements which read tables through a view.
#
#
# 3.1 SELECT statement which uses some table through a view.
#
# Since this statement is not written to the binary log and
# an old version of the table is accessible thanks to how MyISAM
# handles concurrent insert, no locking is necessary.
Success: 'select * from v1' allows concurrent inserts into 't1'.
Success: 'select * from v2' allows concurrent inserts into 't1'.
Success: 'select * from t2 where j in (select i from v1)' allows concurrent inserts into 't1'.
Success: 'select * from t3 where k in (select j from v2)' allows concurrent inserts into 't1'.
#
# 3.2 Statements which modify a table and use views.
#
# Since such statements are going to be written to the binary
# log they need to be serialized against concurrent statements
# and therefore should take strong locks on the data read.
Success: 'update t2 set j= j-10 where j in (select i from v1)' doesn't allow concurrent inserts into 't1'.
Success: 'update t3 set k= k-10 where k in (select j from v2)' doesn't allow concurrent inserts into 't1'.
Success: 'update t2, v1 set j= j-10 where j = i' doesn't allow concurrent inserts into 't1'.
Success: 'update v2 set j= j-10 where j = 3' doesn't allow concurrent inserts into 't1'.
#
# 4. Statements which read tables through stored functions.
#
#
# 4.1 SELECT/SET with a stored function which does not
# modify data and uses SELECT in its turn.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f1()' allows concurrent inserts into 't1'.
Success: 'set @a:= f1()' allows concurrent inserts into 't1'.
#
# 4.2 INSERT (or other statement which modifies data) with
# a stored function which does not modify data and uses
# SELECT.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting the data
# it uses. Therefore it should take strong lock on the data
# it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'insert into t2 values (f1() + 5)' allows concurrent inserts into 't1'.
#
# 4.3 SELECT/SET with a stored function which
# reads and modifies data.
#
# Since a call to such function is written to the binary log,
# it should be serialized with concurrent statements affecting
# the data it uses. Hence, a strong lock on the data read
# should be taken.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'select f2()' allows concurrent inserts into 't1'.
Success: 'set @a:= f2()' allows concurrent inserts into 't1'.
#
# 4.4. SELECT/SET with a stored function which does not
# modify data and reads a table through subselect
# in a control construct.
#
# Again, in theory a call to this function won't get to the
# binary log and thus no strong lock is needed. But in practice
# we don't detect this fact early enough (get_lock_type_for_table())
# to avoid taking a strong lock.
Success: 'select f3()' doesn't allow concurrent inserts into 't1'.
Success: 'set @a:= f3()' doesn't allow concurrent inserts into 't1'.
Success: 'select f4()' doesn't allow concurrent inserts into 't1'.
Success: 'set @a:= f4()' doesn't allow concurrent inserts into 't1'.
#
# 4.5. INSERT (or other statement which modifies data) with
# a stored function which does not modify data and reads
# the table through a subselect in one of its control
# constructs.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting data it
# uses. Therefore it should take a strong lock on the data
# it reads.
Success: 'insert into t2 values (f3() + 5)' doesn't allow concurrent inserts into 't1'.
Success: 'insert into t2 values (f4() + 6)' doesn't allow concurrent inserts into 't1'.
#
# 4.6 SELECT/SET which uses a stored function with
# DML which reads a table via a subquery.
#
# Since call to such function is written to the binary log
# it should be serialized with concurrent statements.
# Hence reads should take a strong lock.
Success: 'select f5()' doesn't allow concurrent inserts into 't1'.
Success: 'set @a:= f5()' doesn't allow concurrent inserts into 't1'.
#
# 4.7 SELECT/SET which uses a stored function which
# doesn't modify data and reads tables through
# a view.
#
# Once again, in theory, calls to such functions won't
# get into the binary log and thus don't need strong
# locks. In practice this fact is discovered
# too late to have any effect.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken
# in case when simple SELECT is used.
Success: 'select f6()' allows concurrent inserts into 't1'.
Success: 'set @a:= f6()' allows concurrent inserts into 't1'.
Success: 'select f7()' doesn't allow concurrent inserts into 't1'.
Success: 'set @a:= f7()' doesn't allow concurrent inserts into 't1'.
#
# 4.8 INSERT which uses stored function which
# doesn't modify data and reads a table
# through a view.
#
# Since such statement is written to the binary log and
# should be serialized with concurrent statements affecting
# the data it uses. Therefore it should take a strong lock on
# the table it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken
# in case when simple SELECT is used.
Success: 'insert into t3 values (f6() + 5)' allows concurrent inserts into 't1'.
Success: 'insert into t3 values (f7() + 5)' doesn't allow concurrent inserts into 't1'.
#
# 4.9 SELECT which uses a stored function which
# modifies data and reads tables through a view.
#
# Since a call to such function is written to the binary log
# it should be serialized with concurrent statements.
# Hence, reads should take strong locks.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken
# in case when simple SELECT is used.
Success: 'select f8()' allows concurrent inserts into 't1'.
Success: 'select f9()' doesn't allow concurrent inserts into 't1'.
#
# 4.10 SELECT which uses a stored function which doesn't modify
# data and reads a table indirectly, by calling another
# function.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f10()' allows concurrent inserts into 't1'.
#
# 4.11 INSERT which uses a stored function which doesn't modify
# data and reads a table indirectly, by calling another
# function.
#
# Since such statement is written to the binary log, it should
# be serialized with concurrent statements affecting the data it
# uses. Therefore it should take strong locks on data it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'insert into t2 values (f10() + 5)' allows concurrent inserts into 't1'.
#
# 4.12 SELECT which uses a stored function which modifies
# data and reads a table indirectly, by calling another
# function.
#
# Since a call to such function is written to the binary log
# it should be serialized from concurrent statements.
# Hence, read should take a strong lock.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'select f11()' allows concurrent inserts into 't1'.
#
# 4.13 SELECT that reads a table through a subquery passed
# as a parameter to a stored function which modifies
# data.
#
# Even though a call to this function is written to the
# binary log, values of its parameters are written as literals.
# So there is no need to acquire strong locks for tables used in
# the subquery.
Success: 'select f12((select i+10 from t1 where i=1))' allows concurrent inserts into 't1'.
#
# 4.14 INSERT that reads a table via a subquery passed
# as a parameter to a stored function which doesn't
# modify data.
#
# Since this statement is written to the binary log it should
# be serialized with concurrent statements affecting the data it
# uses. Therefore it should take strong locks on the data it reads.
Success: 'insert into t2 values (f13((select i+10 from t1 where i=1)))' doesn't allow concurrent inserts into 't1'.
#
# 5. Statements that read tables through stored procedures.
#
#
# 5.1 CALL statement which reads a table via SELECT.
#
# Since neither this statement nor its components are
# written to the binary log, there is no need to take
# strong locks on the data it reads.
Success: 'call p2(@a)' allows concurrent inserts into 't1'.
#
# 5.2 Function that modifies data and uses CALL,
# which reads a table through SELECT.
#
# Since a call to such function is written to the binary
# log, it should be serialized with concurrent statements.
# Hence, in this case reads should take strong locks on data.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'select f14()' allows concurrent inserts into 't1'.
#
# 5.3 SELECT that calls a function that doesn't modify data and
# uses a CALL statement that reads a table via SELECT.
#
# Calls to such functions won't get into the binary log and
# thus don't need to acquire strong locks.
# In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
# used stored functions may lead to broken SBR" strong locks
# are taken (we accepted it as a trade-off for this fix).
Success: 'select f15()' allows concurrent inserts into 't1'.
#
# 5.4 INSERT which calls function which doesn't modify data and
# uses CALL statement which reads table through SELECT.
#
# Since such statement is written to the binary log it should
# be serialized with concurrent statements affecting data it
# uses. Therefore it should take strong locks on data it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'insert into t2 values (f15()+5)' allows concurrent inserts into 't1'.
#
# 6. Statements that use triggers.
#
#
# 6.1 Statement invoking a trigger that reads table via SELECT.
#
# Since this statement is written to the binary log it should
# be serialized with concurrent statements affecting the data
# it uses. Therefore, it should take strong locks on the data
# it reads.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'insert into t4 values (2)' allows concurrent inserts into 't1'.
#
# 6.2 Statement invoking a trigger that reads table through
# a subquery in a control construct.
#
# The above is true for this statement as well.
Success: 'update t4 set l= 2 where l = 1' doesn't allow concurrent inserts into 't1'.
#
# 6.3 Statement invoking a trigger that reads a table through
# a view.
#
# And for this statement.
Success: 'delete from t4 where l = 1' doesn't allow concurrent inserts into 't1'.
#
# 6.4 Statement invoking a trigger that reads a table through
# a stored function.
#
# And for this statement.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'insert into t5 values (2)' allows concurrent inserts into 't1'.
#
# 6.5 Statement invoking a trigger that reads a table through
# stored procedure.
#
# And for this statement.
# But due to bug #53921 "Wrong locks for SELECTs used stored
# functions may lead to broken SBR" weak locks are taken.
Success: 'update t5 set l= 2 where l = 1' allows concurrent inserts into 't1'.
# Clean-up.
drop function f1;
drop function f2;
drop function f3;
drop function f4;
drop function f5;
drop function f6;
drop function f7;
drop function f8;
drop function f9;
drop function f10;
drop function f11;
drop function f12;
drop function f13;
drop function f14;
drop function f15;
drop view v1, v2;
drop procedure p1;
drop procedure p2;
drop table t1, t2, t3, t4, t5, te;
set @@global.concurrent_insert= @old_concurrent_insert;
mysql-test/t/bug39022.test
View file @ 78c6a8ca
...@@ -24,7 +24,7 @@ START TRANSACTION; ...@@ -24,7 +24,7 @@ START TRANSACTION;
connection thread2; connection thread2;
--echo # in thread2 --echo # in thread2
REPLACE INTO t2 VALUES (-17); REPLACE INTO t2 VALUES (-17);
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE;
connection thread1; connection thread1;
--echo # in thread1 --echo # in thread1
...@@ -37,14 +37,14 @@ START TRANSACTION; ...@@ -37,14 +37,14 @@ START TRANSACTION;
REPLACE INTO t1(a,b) VALUES (65,-50); REPLACE INTO t1(a,b) VALUES (65,-50);
REPLACE INTO t2 VALUES (-91); REPLACE INTO t2 VALUES (-91);
send; send;
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); #waits SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE; #waits
connection thread1; connection thread1;
--echo # in thread1 --echo # in thread1
--echo # should not crash --echo # should not crash
--error ER_LOCK_DEADLOCK --error ER_LOCK_DEADLOCK
SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d); #crashes SELECT d FROM t2,t1 WHERE d=(SELECT MAX(a) FROM t1 WHERE t1.a > t2.d) LOCK IN SHARE MODE; #crashes
connection thread2; connection thread2;
--echo # in thread2 --echo # in thread2
......
mysql-test/t/innodb_mysql_lock2.test 0 → 100644
View file @ 78c6a8ca
# This test covers behavior for InnoDB tables.
--source include/have_innodb.inc
# This test requires statement/mixed mode binary logging.
# Row-based mode puts weaker serializability requirements
# so weaker locks are acquired for it.
--source include/have_binlog_format_mixed_or_statement.inc
# Save the initial number of concurrent sessions.
--source include/count_sessions.inc
--echo #
--echo # Test how do we handle locking in various cases when
--echo # we read data from InnoDB tables.
--echo #
--echo # In fact by performing this test we check two things:
--echo # 1) That SQL-layer correctly determine type of thr_lock.c
--echo # lock to be acquired/passed to InnoDB engine.
--echo # 2) That InnoDB engine correctly interprets this lock
--echo # type and takes necessary row locks or does not
--echo # take them if they are not necessary.
--echo #
--echo # This test makes sense only in REPEATABLE-READ mode as
--echo # in SERIALIZABLE mode all statements that read data take
--echo # shared lock on them to enforce its semantics.
select @@session.tx_isolation;
--echo # Prepare playground by creating tables, views,
--echo # routines and triggers used in tests.
connect (con1, localhost, root,,);
connection default;
--disable_warnings
drop table if exists t0, t1, t2, t3, t4, t5, te;
drop view if exists v1, v2;
drop procedure if exists p1;
drop procedure if exists p2;
drop function if exists f1;
drop function if exists f2;
drop function if exists f3;
drop function if exists f4;
drop function if exists f5;
drop function if exists f6;
drop function if exists f7;
drop function if exists f8;
drop function if exists f9;
drop function if exists f10;
drop function if exists f11;
drop function if exists f12;
drop function if exists f13;
drop function if exists f14;
drop function if exists f15;
--enable_warnings
create table t1 (i int primary key) engine=innodb;
insert into t1 values (1), (2), (3), (4), (5);
create table t2 (j int primary key) engine=innodb;
insert into t2 values (1), (2), (3), (4), (5);
create table t3 (k int primary key) engine=innodb;
insert into t3 values (1), (2), (3);
create table t4 (l int primary key) engine=innodb;
insert into t4 values (1);
create table t5 (l int primary key) engine=innodb;
insert into t5 values (1);
create table te(e int primary key);
insert into te values (1);
create view v1 as select i from t1;
create view v2 as select j from t2 where j in (select i from t1);
create procedure p1(k int) insert into t2 values (k);
delimiter |;
create function f1() returns int
begin
declare j int;
select i from t1 where i = 1 into j;
return j;
end|
create function f2() returns int
begin
declare k int;
select i from t1 where i = 1 into k;
insert into t2 values (k + 5);
return 0;
end|
create function f3() returns int
begin
return (select i from t1 where i = 3);
end|
create function f4() returns int
begin
if (select i from t1 where i = 3) then
return 1;
else
return 0;
end if;
end|
create function f5() returns int
begin
insert into t2 values ((select i from t1 where i = 1) + 5);
return 0;
end|
create function f6() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
return k;
end|
create function f7() returns int
begin
declare k int;
select j from v2 where j = 1 into k;
return k;
end|
create function f8() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
insert into t2 values (k+5);
return k;
end|
create function f9() returns int
begin
update v2 set j=j+10 where j=1;
return 1;
end|
create function f10() returns int
begin
return f1();
end|
create function f11() returns int
begin
declare k int;
set k= f1();
insert into t2 values (k+5);
return k;
end|
create function f12(p int) returns int
begin
insert into t2 values (p);
return p;
end|
create function f13(p int) returns int
begin
return p;
end|
create procedure p2(inout p int)
begin
select i from t1 where i = 1 into p;
end|
create function f14() returns int
begin
declare k int;
call p2(k);
insert into t2 values (k+5);
return k;
end|
create function f15() returns int
begin
declare k int;
call p2(k);
return k;
end|
create trigger t4_bi before insert on t4 for each row
begin
declare k int;
select i from t1 where i=1 into k;
set new.l= k+1;
end|
create trigger t4_bu before update on t4 for each row
begin
if (select i from t1 where i=1) then
set new.l= 2;
end if;
end|
--echo # Trigger below uses insertion of duplicate key in 'te'
--echo # table as a way to abort delete operation.
create trigger t4_bd before delete on t4 for each row
begin
if !(select i from v1 where i=1) then
insert into te values (1);
end if;
end|
create trigger t5_bi before insert on t5 for each row
begin
set new.l= f1()+1;
end|
create trigger t5_bu before update on t5 for each row
begin
declare j int;
call p2(j);
set new.l= j + 1;
end|
delimiter ;|
--echo #
--echo # Set common variables to be used by scripts called below.
--echo #
let $con_aux= con1;
let $table= t1;
--echo #
--echo # 1. Statements that read tables and do not use subqueries.
--echo #
--echo #
--echo # 1.1 Simple SELECT statement.
--echo #
--echo # No locks are necessary as this statement won't be written
--echo # to the binary log and InnoDB supports snapshots.
let $statement= select * from t1;
--source include/check_no_row_lock.inc
--echo #
--echo # 1.2 Multi-UPDATE statement.
--echo #
--echo # Has to take shared locks on rows in the table being read as this
--echo # statement will be written to the binary log and therefore should
--echo # be serialized with concurrent statements.
let $statement= update t2, t1 set j= j - 1 where i = j;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 1.3 Multi-DELETE statement.
--echo #
--echo # The above is true for this statement as well.
let $statement= delete t2 from t1, t2 where i = j;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 1.4 DESCRIBE statement.
--echo #
--echo # This statement does not really read data from the
--echo # target table and thus does not take any lock on it.
--echo # We check this for completeness of coverage.
let $statement= describe t1;
--source include/check_no_row_lock.inc
--echo #
--echo # 1.5 SHOW statements.
--echo #
--echo # The above is true for SHOW statements as well.
let $statement= show create table t1;
--source include/check_no_row_lock.inc
let $statement= show keys from t1;
--source include/check_no_row_lock.inc
--echo #
--echo # 2. Statements which read tables through subqueries.
--echo #
--echo #
--echo # 2.1 CALL with a subquery.
--echo #
--echo # A strong lock is not necessary as this statement is not
--echo # written to the binary log as a whole (it is written
--echo # statement-by-statement) and thanks to MVCC we can always get
--echo # versions of rows prior to the update that has locked them.
--echo # But in practice InnoDB does locking reads for all statements
--echo # other than SELECT (unless it is a READ-COMITTED mode or
--echo # innodb_locks_unsafe_for_binlog is ON).
let $statement= call p1((select i + 5 from t1 where i = 1));
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.2 CREATE TABLE with a subquery.
--echo #
--echo # Has to take shared locks on rows in the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent statements.
let $statement= create table t0 engine=innodb select * from t1;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
drop table t0;
let $statement= create table t0 engine=innodb select j from t2 where j in (select i from t1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
drop table t0;
--echo #
--echo # 2.3 DELETE with a subquery.
--echo #
--echo # The above is true for this statement as well.
let $statement= delete from t2 where j in (select i from t1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.4 MULTI-DELETE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= delete t2 from t3, t2 where k = j and j in (select i from t1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.5 DO with a subquery.
--echo #
--echo # In theory should not take row locks as it is not logged.
--echo # In practice InnoDB takes shared row locks.
let $statement= do (select i from t1 where i = 1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.6 INSERT with a subquery.
--echo #
--echo # Has to take shared locks on rows in the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent statements.
let $statement= insert into t2 select i+5 from t1;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= insert into t2 values ((select i+5 from t1 where i = 4));
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.7 LOAD DATA with a subquery.
--echo #
--echo # The above is true for this statement as well.
let $statement= load data infile '../../std_data/rpl_loaddata.dat' into table t2 (@a, @b) set j= @b + (select i from t1 where i = 1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.8 REPLACE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= replace into t2 select i+5 from t1;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= replace into t2 values ((select i+5 from t1 where i = 4));
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.9 SELECT with a subquery.
--echo #
--echo # Locks are not necessary as this statement is not written
--echo # to the binary log and thanks to MVCC we can always get
--echo # versions of rows prior to the update that has locked them.
--echo #
--echo # Also serves as a test case for bug #46947 "Embedded SELECT
--echo # without FOR UPDATE is causing a lock".
let $statement= select * from t2 where j in (select i from t1);
--source include/check_no_row_lock.inc
--echo #
--echo # 2.10 SET with a subquery.
--echo #
--echo # In theory should not require locking as it is not written
--echo # to the binary log. In practice InnoDB acquires shared row
--echo # locks.
let $statement= set @a:= (select i from t1 where i = 1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.11 SHOW with a subquery.
--echo #
--echo # Similarly to the previous case, in theory should not require locking
--echo # as it is not written to the binary log. In practice InnoDB
--echo # acquires shared row locks.
let $statement= show tables from test where Tables_in_test = 't2' and (select i from t1 where i = 1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= show columns from t2 where (select i from t1 where i = 1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.12 UPDATE with a subquery.
--echo #
--echo # Has to take shared locks on rows in the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent statements.
let $statement= update t2 set j= j-10 where j in (select i from t1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 2.13 MULTI-UPDATE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= update t2, t3 set j= j -10 where j=k and j in (select i from t1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 3. Statements which read tables through a view.
--echo #
--echo #
--echo # 3.1 SELECT statement which uses some table through a view.
--echo #
--echo # Since this statement is not written to the binary log
--echo # and old version of rows are accessible thanks to MVCC,
--echo # no locking is necessary.
let $statement= select * from v1;
--source include/check_no_row_lock.inc
let $statement= select * from v2;
--source include/check_no_row_lock.inc
let $statement= select * from t2 where j in (select i from v1);
--source include/check_no_row_lock.inc
let $statement= select * from t3 where k in (select j from v2);
--source include/check_no_row_lock.inc
--echo #
--echo # 3.2 Statements which modify a table and use views.
--echo #
--echo # Since such statements are going to be written to the binary
--echo # log they need to be serialized against concurrent statements
--echo # and therefore should take shared row locks on data read.
let $statement= update t2 set j= j-10 where j in (select i from v1);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= update t3 set k= k-10 where k in (select j from v2);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= update t2, v1 set j= j-10 where j = i;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= update v2 set j= j-10 where j = 3;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4. Statements which read tables through stored functions.
--echo #
--echo #
--echo # 4.1 SELECT/SET with a stored function which does not
--echo # modify data and uses SELECT in its turn.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f1();
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
let $statement= set @a:= f1();
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.2 INSERT (or other statement which modifies data) with
--echo # a stored function which does not modify data and uses
--echo # SELECT.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data
--echo # it uses. Therefore it should take row locks on the data
--echo # it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= insert into t2 values (f1() + 5);
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.3 SELECT/SET with a stored function which
--echo # reads and modifies data.
--echo #
--echo # Since a call to such function is written to the binary log,
--echo # it should be serialized with concurrent statements affecting
--echo # the data it uses. Hence, row locks on the data read
--echo # should be taken.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= select f2();
let $wait_statement= select i from t1 where i = 1 into k;
--source include/check_no_row_lock.inc
let $statement= set @a:= f2();
let $wait_statement= select i from t1 where i = 1 into k;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.4. SELECT/SET with a stored function which does not
--echo # modify data and reads a table through subselect
--echo # in a control construct.
--echo #
--echo # Again, in theory a call to this function won't get to the
--echo # binary log and thus no locking is needed. But in practice
--echo # we don't detect this fact early enough (get_lock_type_for_table())
--echo # to avoid taking row locks.
let $statement= select f3();
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= set @a:= f3();
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= select f4();
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= set @a:= f4();
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.5. INSERT (or other statement which modifies data) with
--echo # a stored function which does not modify data and reads
--echo # the table through a subselect in one of its control
--echo # constructs.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting data it
--echo # uses. Therefore it should take row locks on the data
--echo # it reads.
let $statement= insert into t2 values (f3() + 5);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
let $statement= insert into t2 values (f4() + 6);
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.6 SELECT/SET which uses a stored function with
--echo # DML which reads a table via a subquery.
--echo #
--echo # Since call to such function is written to the binary log
--echo # it should be serialized with concurrent statements.
--echo # Hence reads should take row locks.
let $statement= select f5();
let $wait_statement= insert into t2 values ((select i from t1 where i = 1) + 5);
--source include/check_shared_row_lock.inc
let $statement= set @a:= f5();
let $wait_statement= insert into t2 values ((select i from t1 where i = 1) + 5);
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.7 SELECT/SET which uses a stored function which
--echo # doesn't modify data and reads tables through
--echo # a view.
--echo #
--echo # Once again, in theory, calls to such functions won't
--echo # get into the binary log and thus don't need row
--echo # locks. In practice this fact is discovered
--echo # too late to have any effect.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken
--echo # in case of simple SELECT.
let $statement= select f6();
let $wait_statement= select i from v1 where i = 1 into k;
--source include/check_no_row_lock.inc
let $statement= set @a:= f6();
let $wait_statement= select i from v1 where i = 1 into k;
--source include/check_no_row_lock.inc
let $statement= select f7();
let $wait_statement= select j from v2 where j = 1 into k;
--source include/check_shared_row_lock.inc
let $statement= set @a:= f7();
let $wait_statement= select j from v2 where j = 1 into k;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.8 INSERT which uses stored function which
--echo # doesn't modify data and reads a table
--echo # through a view.
--echo #
--echo # Since such statement is written to the binary log and
--echo # should be serialized with concurrent statements affecting
--echo # the data it uses. Therefore it should take row locks on
--echo # the rows it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken
--echo # in case of simple SELECT.
let $statement= insert into t3 values (f6() + 5);
let $wait_statement= select i from v1 where i = 1 into k;
--source include/check_no_row_lock.inc
let $statement= insert into t3 values (f7() + 5);
let $wait_statement= select j from v2 where j = 1 into k;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.9 SELECT which uses a stored function which
--echo # modifies data and reads tables through a view.
--echo #
--echo # Since a call to such function is written to the binary log
--echo # it should be serialized with concurrent statements.
--echo # Hence, reads should take row locks.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken
--echo # in case of simple SELECT.
let $statement= select f8();
let $wait_statement= select i from v1 where i = 1 into k;
--source include/check_no_row_lock.inc
let $statement= select f9();
let $wait_statement= update v2 set j=j+10 where j=1;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.10 SELECT which uses stored function which doesn't modify
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f10();
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.11 INSERT which uses a stored function which doesn't modify
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Since such statement is written to the binary log, it should
--echo # be serialized with concurrent statements affecting the data it
--echo # uses. Therefore it should take row locks on data it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= insert into t2 values (f10() + 5);
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.12 SELECT which uses a stored function which modifies
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Since a call to such function is written to the binary log
--echo # it should be serialized from concurrent statements.
--echo # Hence, reads should take row locks.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= select f11();
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 4.13 SELECT that reads a table through a subquery passed
--echo # as a parameter to a stored function which modifies
--echo # data.
--echo #
--echo # Even though a call to this function is written to the
--echo # binary log, values of its parameters are written as literals.
--echo # So there is no need to acquire row locks on rows used in
--echo # the subquery.
--echo # But due to the fact that in 5.1 for prelocked statements
--echo # THD::in_lock_tables is set to TRUE we acquire strong locks
--echo # (see also bug#44613 "SELECT statement inside FUNCTION takes
--echo # a shared lock" [sic!!!]).
let $statement= select f12((select i+10 from t1 where i=1));
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 4.14 INSERT that reads a table via a subquery passed
--echo # as a parameter to a stored function which doesn't
--echo # modify data.
--echo #
--echo # Since this statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data it
--echo # uses. Therefore it should take row locks on the data it reads.
let $statement= insert into t2 values (f13((select i+10 from t1 where i=1)));
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 5. Statements that read tables through stored procedures.
--echo #
--echo #
--echo # 5.1 CALL statement which reads a table via SELECT.
--echo #
--echo # Since neither this statement nor its components are
--echo # written to the binary log, there is no need to take
--echo # row locks on the data it reads.
let $statement= call p2(@a);
--source include/check_no_row_lock.inc
--echo #
--echo # 5.2 Function that modifies data and uses CALL,
--echo # which reads a table through SELECT.
--echo #
--echo # Since a call to such function is written to the binary
--echo # log, it should be serialized with concurrent statements.
--echo # Hence, in this case reads should take row locks on data.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= select f14();
let $wait_statement= select i from t1 where i = 1 into p;
--source include/check_no_row_lock.inc
--echo #
--echo # 5.3 SELECT that calls a function that doesn't modify data and
--echo # uses a CALL statement that reads a table via SELECT.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f15();
let $wait_statement= select i from t1 where i = 1 into p;
--source include/check_no_row_lock.inc
--echo #
--echo # 5.4 INSERT which calls function which doesn't modify data and
--echo # uses CALL statement which reads table through SELECT.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting data it
--echo # uses. Therefore it should take row locks on data it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= insert into t2 values (f15()+5);
let $wait_statement= select i from t1 where i = 1 into p;
--source include/check_no_row_lock.inc
--echo #
--echo # 6. Statements that use triggers.
--echo #
--echo #
--echo # 6.1 Statement invoking a trigger that reads table via SELECT.
--echo #
--echo # Since this statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data
--echo # it uses. Therefore, it should take row locks on the data
--echo # it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= insert into t4 values (2);
let $wait_statement= select i from t1 where i=1 into k;
--source include/check_no_row_lock.inc
--echo #
--echo # 6.2 Statement invoking a trigger that reads table through
--echo # a subquery in a control construct.
--echo #
--echo # The above is true for this statement as well.
let $statement= update t4 set l= 2 where l = 1;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 6.3 Statement invoking a trigger that reads a table through
--echo # a view.
--echo #
--echo # And for this statement.
let $statement= delete from t4 where l = 1;
let $wait_statement= $statement;
--source include/check_shared_row_lock.inc
--echo #
--echo # 6.4 Statement invoking a trigger that reads a table through
--echo # a stored function.
--echo #
--echo # And for this statement.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= insert into t5 values (2);
let $wait_statement= select i from t1 where i = 1 into j;
--source include/check_no_row_lock.inc
--echo #
--echo # 6.5 Statement invoking a trigger that reads a table through
--echo # stored procedure.
--echo #
--echo # And for this statement.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" no lock is taken.
let $statement= update t5 set l= 2 where l = 1;
let $wait_statement= select i from t1 where i = 1 into p;
--source include/check_no_row_lock.inc
--echo # Clean-up.
drop function f1;
drop function f2;
drop function f3;
drop function f4;
drop function f5;
drop function f6;
drop function f7;
drop function f8;
drop function f9;
drop function f10;
drop function f11;
drop function f12;
drop function f13;
drop function f14;
drop function f15;
drop view v1, v2;
drop procedure p1;
drop procedure p2;
drop table t1, t2, t3, t4, t5, te;
disconnect con1;
# Check that all connections opened by test cases in this file are really
# gone so execution of other tests won't be affected by their presence.
--source include/wait_until_count_sessions.inc
mysql-test/t/lock_sync.test 0 → 100644
View file @ 78c6a8ca
#
# Locking related tests which use DEBUG_SYNC facility.
#
--source include/have_debug_sync.inc
# This test requires statement/mixed mode binary logging.
# Row-based mode puts weaker serializability requirements
# so weaker locks are acquired for it.
--source include/have_binlog_format_mixed_or_statement.inc
# Save the initial number of concurrent sessions.
--source include/count_sessions.inc
--echo #
--echo # Test how we handle locking in various cases when
--echo # we read data from MyISAM tables.
--echo #
--echo # In this test we mostly check that the SQL-layer correctly
--echo # determines the type of thr_lock.c lock for a table being
--echo # read.
--echo # I.e. that it disallows concurrent inserts when the statement
--echo # is going to be written to the binary log and therefore
--echo # should be serialized, and allows concurrent inserts when
--echo # such serialization is not necessary (e.g. when
--echo # the statement is not written to binary log).
--echo #
--echo # Force concurrent inserts to be performed even if the table
--echo # has gaps. This allows to simplify clean up in scripts
--echo # used below (instead of backing up table being inserted
--echo # into and then restoring it from backup at the end of the
--echo # script we can simply delete rows which were inserted).
set @old_concurrent_insert= @@global.concurrent_insert;
set @@global.concurrent_insert= 2;
select @@global.concurrent_insert;
--echo # Prepare playground by creating tables, views,
--echo # routines and triggers used in tests.
connect (con1, localhost, root,,);
connect (con2, localhost, root,,);
connection default;
--disable_warnings
drop table if exists t0, t1, t2, t3, t4, t5, te;
drop view if exists v1, v2;
drop procedure if exists p1;
drop procedure if exists p2;
drop function if exists f1;
drop function if exists f2;
drop function if exists f3;
drop function if exists f4;
drop function if exists f5;
drop function if exists f6;
drop function if exists f7;
drop function if exists f8;
drop function if exists f9;
drop function if exists f10;
drop function if exists f11;
drop function if exists f12;
drop function if exists f13;
drop function if exists f14;
drop function if exists f15;
--enable_warnings
create table t1 (i int primary key);
insert into t1 values (1), (2), (3), (4), (5);
create table t2 (j int primary key);
insert into t2 values (1), (2), (3), (4), (5);
create table t3 (k int primary key);
insert into t3 values (1), (2), (3);
create table t4 (l int primary key);
insert into t4 values (1);
create table t5 (l int primary key);
insert into t5 values (1);
create table te(e int primary key);
insert into te values (1);
create view v1 as select i from t1;
create view v2 as select j from t2 where j in (select i from t1);
create procedure p1(k int) insert into t2 values (k);
delimiter |;
create function f1() returns int
begin
declare j int;
select i from t1 where i = 1 into j;
return j;
end|
create function f2() returns int
begin
declare k int;
select i from t1 where i = 1 into k;
insert into t2 values (k + 5);
return 0;
end|
create function f3() returns int
begin
return (select i from t1 where i = 3);
end|
create function f4() returns int
begin
if (select i from t1 where i = 3) then
return 1;
else
return 0;
end if;
end|
create function f5() returns int
begin
insert into t2 values ((select i from t1 where i = 1) + 5);
return 0;
end|
create function f6() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
return k;
end|
create function f7() returns int
begin
declare k int;
select j from v2 where j = 1 into k;
return k;
end|
create function f8() returns int
begin
declare k int;
select i from v1 where i = 1 into k;
insert into t2 values (k+5);
return k;
end|
create function f9() returns int
begin
update v2 set j=j+10 where j=1;
return 1;
end|
create function f10() returns int
begin
return f1();
end|
create function f11() returns int
begin
declare k int;
set k= f1();
insert into t2 values (k+5);
return k;
end|
create function f12(p int) returns int
begin
insert into t2 values (p);
return p;
end|
create function f13(p int) returns int
begin
return p;
end|
create procedure p2(inout p int)
begin
select i from t1 where i = 1 into p;
end|
create function f14() returns int
begin
declare k int;
call p2(k);
insert into t2 values (k+5);
return k;
end|
create function f15() returns int
begin
declare k int;
call p2(k);
return k;
end|
create trigger t4_bi before insert on t4 for each row
begin
declare k int;
select i from t1 where i=1 into k;
set new.l= k+1;
end|
create trigger t4_bu before update on t4 for each row
begin
if (select i from t1 where i=1) then
set new.l= 2;
end if;
end|
--echo # Trigger below uses insertion of duplicate key in 'te'
--echo # table as a way to abort delete operation.
create trigger t4_bd before delete on t4 for each row
begin
if !(select i from v1 where i=1) then
insert into te values (1);
end if;
end|
create trigger t5_bi before insert on t5 for each row
begin
set new.l= f1()+1;
end|
create trigger t5_bu before update on t5 for each row
begin
declare j int;
call p2(j);
set new.l= j + 1;
end|
delimiter ;|
--echo #
--echo # Set common variables to be used by the scripts
--echo # called below.
--echo #
let $con_aux1= con1;
let $con_aux2= con2;
let $table= t1;
--echo # Switch to connection 'con1'.
connection con1;
--echo # Cache all functions used in the tests below so statements
--echo # calling them won't need to open and lock mysql.proc table
--echo # and we can assume that each statement locks its tables
--echo # once during its execution.
--disable_result_log
show create procedure p1;
show create procedure p2;
show create function f1;
show create function f2;
show create function f3;
show create function f4;
show create function f5;
show create function f6;
show create function f7;
show create function f8;
show create function f9;
show create function f10;
show create function f11;
show create function f12;
show create function f13;
show create function f14;
show create function f15;
--enable_result_log
--echo # Switch back to connection 'default'.
connection default;
--echo #
--echo # 1. Statements that read tables and do not use subqueries.
--echo #
--echo #
--echo # 1.1 Simple SELECT statement.
--echo #
--echo # No locks are necessary as this statement won't be written
--echo # to the binary log and thanks to how MyISAM works SELECT
--echo # will see version of the table prior to concurrent insert.
let $statement= select * from t1;
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 1.2 Multi-UPDATE statement.
--echo #
--echo # Has to take shared locks on rows in the table being read as this
--echo # statement will be written to the binary log and therefore should
--echo # be serialized with concurrent statements.
let $statement= update t2, t1 set j= j - 1 where i = j;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 1.3 Multi-DELETE statement.
--echo #
--echo # The above is true for this statement as well.
let $statement= delete t2 from t1, t2 where i = j;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 1.4 DESCRIBE statement.
--echo #
--echo # This statement does not really read data from the
--echo # target table and thus does not take any lock on it.
--echo # We check this for completeness of coverage.
lock table t1 write;
--echo # Switching to connection 'con1'.
connection con1;
--echo # This statement should not be blocked.
--disable_result_log
describe t1;
--enable_result_log
--echo # Switching to connection 'default'.
connection default;
unlock tables;
--echo #
--echo # 1.5 SHOW statements.
--echo #
--echo # The above is true for SHOW statements as well.
lock table t1 write;
--echo # Switching to connection 'con1'.
connection con1;
--echo # These statements should not be blocked.
# The below test for SHOW CREATE TABLE is disabled until bug 52593
# "SHOW CREATE TABLE is blocked if table is locked for write by another
# connection" is fixed.
--disable_parsing
show create table t1;
--enable_parsing
--disable_result_log
show keys from t1;
--enable_result_log
--echo # Switching to connection 'default'.
connection default;
unlock tables;
--echo #
--echo # 2. Statements which read tables through subqueries.
--echo #
--echo #
--echo # 2.1 CALL with a subquery.
--echo #
--echo # In theory strong lock is not necessary as this statement
--echo # is not written to the binary log as a whole (it is written
--echo # statement-by-statement). But in practice in 5.1 for
--echo # almost everything except SELECT we take strong lock.
let $statement= call p1((select i + 5 from t1 where i = 1));
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.2 CREATE TABLE with a subquery.
--echo #
--echo # Has to take a strong lock on the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent statements.
let $statement= create table t0 select * from t1;
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
drop table t0;
let $statement= create table t0 select j from t2 where j in (select i from t1);
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
drop table t0;
--echo #
--echo # 2.3 DELETE with a subquery.
--echo #
--echo # The above is true for this statement as well.
let $statement= delete from t2 where j in (select i from t1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.4 MULTI-DELETE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= delete t2 from t3, t2 where k = j and j in (select i from t1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.5 DO with a subquery.
--echo #
--echo # In theory strong lock is not necessary as it is not logged.
--echo # But in practice in 5.1 for almost everything except SELECT
--echo # we take strong lock.
let $statement= do (select i from t1 where i = 1);
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.6 INSERT with a subquery.
--echo #
--echo # Has to take a strong lock on the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent inserts.
let $statement= insert into t2 select i+5 from t1;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= insert into t2 values ((select i+5 from t1 where i = 4));
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.7 LOAD DATA with a subquery.
--echo #
--echo # The above is true for this statement as well.
let $statement= load data infile '../../std_data/rpl_loaddata.dat' into table t2 (@a, @b) set j= @b + (select i from t1 where i = 1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.8 REPLACE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= replace into t2 select i+5 from t1;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= replace into t2 values ((select i+5 from t1 where i = 4));
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.9 SELECT with a subquery.
--echo #
--echo # Strong locks are not necessary as this statement is not written
--echo # to the binary log and thanks to how MyISAM works this statement
--echo # sees a version of the table prior to the concurrent insert.
let $statement= select * from t2 where j in (select i from t1);
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 2.10 SET with a subquery.
--echo #
--echo # In theory the same is true for this statement as well.
--echo # But in practice in 5.1 we acquire strong lock in this
--echo # case as well.
let $statement= set @a:= (select i from t1 where i = 1);
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.11 SHOW with a subquery.
--echo #
--echo # The same is true for this statement too.
let $statement= show tables from test where Tables_in_test = 't2' and (select i from t1 where i = 1);
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
let $statement= show columns from t2 where (select i from t1 where i = 1);
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.12 UPDATE with a subquery.
--echo #
--echo # Has to take a strong lock on the table being read as
--echo # this statement is written to the binary log and therefore
--echo # should be serialized with concurrent inserts.
let $statement= update t2 set j= j-10 where j in (select i from t1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 2.13 MULTI-UPDATE with a subquery.
--echo #
--echo # Same is true for this statement as well.
let $statement= update t2, t3 set j= j -10 where j=k and j in (select i from t1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 3. Statements which read tables through a view.
--echo #
--echo #
--echo # 3.1 SELECT statement which uses some table through a view.
--echo #
--echo # Since this statement is not written to the binary log and
--echo # an old version of the table is accessible thanks to how MyISAM
--echo # handles concurrent insert, no locking is necessary.
let $statement= select * from v1;
let $restore_table= ;
--source include/check_concurrent_insert.inc
let $statement= select * from v2;
let $restore_table= ;
--source include/check_concurrent_insert.inc
let $statement= select * from t2 where j in (select i from v1);
let $restore_table= ;
--source include/check_concurrent_insert.inc
let $statement= select * from t3 where k in (select j from v2);
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 3.2 Statements which modify a table and use views.
--echo #
--echo # Since such statements are going to be written to the binary
--echo # log they need to be serialized against concurrent statements
--echo # and therefore should take strong locks on the data read.
let $statement= update t2 set j= j-10 where j in (select i from v1);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= update t3 set k= k-10 where k in (select j from v2);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= update t2, v1 set j= j-10 where j = i;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= update v2 set j= j-10 where j = 3;
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4. Statements which read tables through stored functions.
--echo #
--echo #
--echo # 4.1 SELECT/SET with a stored function which does not
--echo # modify data and uses SELECT in its turn.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f1();
let $restore_table= ;
--source include/check_concurrent_insert.inc
let $statement= set @a:= f1();
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.2 INSERT (or other statement which modifies data) with
--echo # a stored function which does not modify data and uses
--echo # SELECT.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data
--echo # it uses. Therefore it should take strong lock on the data
--echo # it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= insert into t2 values (f1() + 5);
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.3 SELECT/SET with a stored function which
--echo # reads and modifies data.
--echo #
--echo # Since a call to such function is written to the binary log,
--echo # it should be serialized with concurrent statements affecting
--echo # the data it uses. Hence, a strong lock on the data read
--echo # should be taken.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= select f2();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
let $statement= set @a:= f2();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.4. SELECT/SET with a stored function which does not
--echo # modify data and reads a table through subselect
--echo # in a control construct.
--echo #
--echo # Again, in theory a call to this function won't get to the
--echo # binary log and thus no strong lock is needed. But in practice
--echo # we don't detect this fact early enough (get_lock_type_for_table())
--echo # to avoid taking a strong lock.
let $statement= select f3();
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
let $statement= set @a:= f3();
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
let $statement= select f4();
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
let $statement= set @a:= f4();
let $restore_table= ;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.5. INSERT (or other statement which modifies data) with
--echo # a stored function which does not modify data and reads
--echo # the table through a subselect in one of its control
--echo # constructs.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting data it
--echo # uses. Therefore it should take a strong lock on the data
--echo # it reads.
let $statement= insert into t2 values (f3() + 5);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= insert into t2 values (f4() + 6);
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.6 SELECT/SET which uses a stored function with
--echo # DML which reads a table via a subquery.
--echo #
--echo # Since call to such function is written to the binary log
--echo # it should be serialized with concurrent statements.
--echo # Hence reads should take a strong lock.
let $statement= select f5();
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= set @a:= f5();
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.7 SELECT/SET which uses a stored function which
--echo # doesn't modify data and reads tables through
--echo # a view.
--echo #
--echo # Once again, in theory, calls to such functions won't
--echo # get into the binary log and thus don't need strong
--echo # locks. In practice this fact is discovered
--echo # too late to have any effect.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken
--echo # in case when simple SELECT is used.
let $statement= select f6();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
let $statement= set @a:= f6();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
let $statement= select f7();
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
let $statement= set @a:= f7();
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.8 INSERT which uses stored function which
--echo # doesn't modify data and reads a table
--echo # through a view.
--echo #
--echo # Since such statement is written to the binary log and
--echo # should be serialized with concurrent statements affecting
--echo # the data it uses. Therefore it should take a strong lock on
--echo # the table it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken
--echo # in case when simple SELECT is used.
let $statement= insert into t3 values (f6() + 5);
let $restore_table= t3;
--source include/check_concurrent_insert.inc
let $statement= insert into t3 values (f7() + 5);
let $restore_table= t3;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.9 SELECT which uses a stored function which
--echo # modifies data and reads tables through a view.
--echo #
--echo # Since a call to such function is written to the binary log
--echo # it should be serialized with concurrent statements.
--echo # Hence, reads should take strong locks.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken
--echo # in case when simple SELECT is used.
let $statement= select f8();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
let $statement= select f9();
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 4.10 SELECT which uses a stored function which doesn't modify
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f10();
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.11 INSERT which uses a stored function which doesn't modify
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Since such statement is written to the binary log, it should
--echo # be serialized with concurrent statements affecting the data it
--echo # uses. Therefore it should take strong locks on data it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= insert into t2 values (f10() + 5);
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.12 SELECT which uses a stored function which modifies
--echo # data and reads a table indirectly, by calling another
--echo # function.
--echo #
--echo # Since a call to such function is written to the binary log
--echo # it should be serialized from concurrent statements.
--echo # Hence, read should take a strong lock.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= select f11();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.13 SELECT that reads a table through a subquery passed
--echo # as a parameter to a stored function which modifies
--echo # data.
--echo #
--echo # Even though a call to this function is written to the
--echo # binary log, values of its parameters are written as literals.
--echo # So there is no need to acquire strong locks for tables used in
--echo # the subquery.
let $statement= select f12((select i+10 from t1 where i=1));
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 4.14 INSERT that reads a table via a subquery passed
--echo # as a parameter to a stored function which doesn't
--echo # modify data.
--echo #
--echo # Since this statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data it
--echo # uses. Therefore it should take strong locks on the data it reads.
let $statement= insert into t2 values (f13((select i+10 from t1 where i=1)));
let $restore_table= t2;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 5. Statements that read tables through stored procedures.
--echo #
--echo #
--echo # 5.1 CALL statement which reads a table via SELECT.
--echo #
--echo # Since neither this statement nor its components are
--echo # written to the binary log, there is no need to take
--echo # strong locks on the data it reads.
let $statement= call p2(@a);
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 5.2 Function that modifies data and uses CALL,
--echo # which reads a table through SELECT.
--echo #
--echo # Since a call to such function is written to the binary
--echo # log, it should be serialized with concurrent statements.
--echo # Hence, in this case reads should take strong locks on data.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= select f14();
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 5.3 SELECT that calls a function that doesn't modify data and
--echo # uses a CALL statement that reads a table via SELECT.
--echo #
--echo # Calls to such functions won't get into the binary log and
--echo # thus don't need to acquire strong locks.
--echo # In 5.5 due to fix for bug #53921 "Wrong locks for SELECTs
--echo # used stored functions may lead to broken SBR" strong locks
--echo # are taken (we accepted it as a trade-off for this fix).
let $statement= select f15();
let $restore_table= ;
--source include/check_concurrent_insert.inc
--echo #
--echo # 5.4 INSERT which calls function which doesn't modify data and
--echo # uses CALL statement which reads table through SELECT.
--echo #
--echo # Since such statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting data it
--echo # uses. Therefore it should take strong locks on data it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= insert into t2 values (f15()+5);
let $restore_table= t2;
--source include/check_concurrent_insert.inc
--echo #
--echo # 6. Statements that use triggers.
--echo #
--echo #
--echo # 6.1 Statement invoking a trigger that reads table via SELECT.
--echo #
--echo # Since this statement is written to the binary log it should
--echo # be serialized with concurrent statements affecting the data
--echo # it uses. Therefore, it should take strong locks on the data
--echo # it reads.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= insert into t4 values (2);
let $restore_table= t4;
--source include/check_concurrent_insert.inc
--echo #
--echo # 6.2 Statement invoking a trigger that reads table through
--echo # a subquery in a control construct.
--echo #
--echo # The above is true for this statement as well.
let $statement= update t4 set l= 2 where l = 1;
let $restore_table= t4;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 6.3 Statement invoking a trigger that reads a table through
--echo # a view.
--echo #
--echo # And for this statement.
let $statement= delete from t4 where l = 1;
let $restore_table= t4;
--source include/check_no_concurrent_insert.inc
--echo #
--echo # 6.4 Statement invoking a trigger that reads a table through
--echo # a stored function.
--echo #
--echo # And for this statement.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= insert into t5 values (2);
let $restore_table= t5;
--source include/check_concurrent_insert.inc
--echo #
--echo # 6.5 Statement invoking a trigger that reads a table through
--echo # stored procedure.
--echo #
--echo # And for this statement.
--echo # But due to bug #53921 "Wrong locks for SELECTs used stored
--echo # functions may lead to broken SBR" weak locks are taken.
let $statement= update t5 set l= 2 where l = 1;
let $restore_table= t5;
--source include/check_concurrent_insert.inc
--echo # Clean-up.
drop function f1;
drop function f2;
drop function f3;
drop function f4;
drop function f5;
drop function f6;
drop function f7;
drop function f8;
drop function f9;
drop function f10;
drop function f11;
drop function f12;
drop function f13;
drop function f14;
drop function f15;
drop view v1, v2;
drop procedure p1;
drop procedure p2;
drop table t1, t2, t3, t4, t5, te;
disconnect con1;
disconnect con2;
set @@global.concurrent_insert= @old_concurrent_insert;
# Check that all connections opened by test cases in this file are really
# gone so execution of other tests won't be affected by their presence.
--source include/wait_until_count_sessions.inc
sql/mysql_priv.h
View file @ 78c6a8ca
...@@ -1288,7 +1288,8 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last, ...@@ -1288,7 +1288,8 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last,
TABLE_LIST *new_child_list, TABLE_LIST **new_last); TABLE_LIST *new_child_list, TABLE_LIST **new_last);
bool reopen_table(TABLE *table); bool reopen_table(TABLE *table);
bool reopen_tables(THD *thd,bool get_locks,bool in_refresh); bool reopen_tables(THD *thd,bool get_locks,bool in_refresh);
thr_lock_type read_lock_type_for_table(THD *thd, TABLE *table); thr_lock_type read_lock_type_for_table(THD *thd, LEX *lex,
TABLE_LIST *table_list);
void close_data_files_and_morph_locks(THD *thd, const char *db, void close_data_files_and_morph_locks(THD *thd, const char *db,
const char *table_name); const char *table_name);
void close_handle_and_leave_table_as_lock(TABLE *table); void close_handle_and_leave_table_as_lock(TABLE *table);
......
sql/sql_base.cc
View file @ 78c6a8ca
...@@ -4418,7 +4418,8 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last, ...@@ -4418,7 +4418,8 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last,
Return a appropriate read lock type given a table object. Return a appropriate read lock type given a table object.
@param thd Thread context @param thd Thread context
@param table TABLE object for table to be locked @param lex LEX for the current statement.
@param table_list Table list element for table to be locked.
@remark Due to a statement-based replication limitation, statements such as @remark Due to a statement-based replication limitation, statements such as
INSERT INTO .. SELECT FROM .. and CREATE TABLE .. SELECT FROM need INSERT INTO .. SELECT FROM .. and CREATE TABLE .. SELECT FROM need
...@@ -4427,19 +4428,32 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last, ...@@ -4427,19 +4428,32 @@ bool fix_merge_after_open(TABLE_LIST *old_child_list, TABLE_LIST **old_last,
source table. If such a statement gets applied on the slave before source table. If such a statement gets applied on the slave before
the INSERT .. SELECT statement finishes, data on the master could the INSERT .. SELECT statement finishes, data on the master could
differ from data on the slave and end-up with a discrepancy between differ from data on the slave and end-up with a discrepancy between
the binary log and table state. Furthermore, this does not apply to the binary log and table state.
I_S and log tables as it's always unsafe to replicate such tables This also applies to SELECT/SET/DO statements which use stored
under statement-based replication as the table on the slave might functions. Calls to such functions are going to be logged as a
contain other data (ie: general_log is enabled on the slave). The whole and thus should be serialized against concurrent changes
statement will be marked as unsafe for SBR in decide_logging_format(). to tables used by those functions. This can be avoided if functions
only read data but doing so requires more complex analysis than it
is done now (unfortunately, due to bug #53921 "Wrong locks for
SELECTs used stored functions may lead to broken SBR" this rule
is not followed in cases when stored function or trigger use
simple SELECT and not a subselect in their body).
Furthermore, this does not apply to I_S and log tables as it's
always unsafe to replicate such tables under statement-based
replication as the table on the slave might contain other data
(ie: general_log is enabled on the slave). The statement will
be marked as unsafe for SBR in decide_logging_format().
*/ */
thr_lock_type read_lock_type_for_table(THD *thd, TABLE *table) thr_lock_type read_lock_type_for_table(THD *thd, LEX *lex,
TABLE_LIST *table_list)
{ {
bool log_on= mysql_bin_log.is_open() && (thd->options & OPTION_BIN_LOG); bool log_on= mysql_bin_log.is_open() && (thd->options & OPTION_BIN_LOG);
ulong binlog_format= thd->variables.binlog_format; ulong binlog_format= thd->variables.binlog_format;
if ((log_on == FALSE) || (binlog_format == BINLOG_FORMAT_ROW) || if ((log_on == FALSE) || (binlog_format == BINLOG_FORMAT_ROW) ||
(table->s->table_category == TABLE_CATEGORY_PERFORMANCE)) (table_list->table->s->table_category == TABLE_CATEGORY_PERFORMANCE) ||
(lex->sql_command == SQLCOM_SELECT &&
! table_list->prelocking_placeholder))
return TL_READ; return TL_READ;
else else
return TL_READ_NO_INSERT; return TL_READ_NO_INSERT;
...@@ -4735,7 +4749,7 @@ int open_tables(THD *thd, TABLE_LIST **start, uint *counter, uint flags) ...@@ -4735,7 +4749,7 @@ int open_tables(THD *thd, TABLE_LIST **start, uint *counter, uint flags)
tables->table->reginfo.lock_type= thd->update_lock_default; tables->table->reginfo.lock_type= thd->update_lock_default;
else if (tables->lock_type == TL_READ_DEFAULT) else if (tables->lock_type == TL_READ_DEFAULT)
tables->table->reginfo.lock_type= tables->table->reginfo.lock_type=
read_lock_type_for_table(thd, tables->table); read_lock_type_for_table(thd, thd->lex, tables);
else else
tables->table->reginfo.lock_type= tables->lock_type; tables->table->reginfo.lock_type= tables->lock_type;
} }
...@@ -5389,6 +5403,8 @@ int lock_tables(THD *thd, TABLE_LIST *tables, uint count, bool *need_reopen) ...@@ -5389,6 +5403,8 @@ int lock_tables(THD *thd, TABLE_LIST *tables, uint count, bool *need_reopen)
DBUG_RETURN(-1); DBUG_RETURN(-1);
} }
DEBUG_SYNC(thd, "after_lock_tables_takes_lock");
if (thd->lex->requires_prelocking() && if (thd->lex->requires_prelocking() &&
thd->lex->sql_command != SQLCOM_LOCK_TABLES) thd->lex->sql_command != SQLCOM_LOCK_TABLES)
{ {
......
sql/sql_update.cc
View file @ 78c6a8ca
...@@ -1053,7 +1053,7 @@ int mysql_multi_update_prepare(THD *thd) ...@@ -1053,7 +1053,7 @@ int mysql_multi_update_prepare(THD *thd)
correct order of statements. Otherwise, we use a TL_READ lock to correct order of statements. Otherwise, we use a TL_READ lock to
improve performance. improve performance.
*/ */
tl->lock_type= read_lock_type_for_table(thd, table); tl->lock_type= read_lock_type_for_table(thd, lex, tl);
tl->updating= 0; tl->updating= 0;
/* Update TABLE::lock_type accordingly. */ /* Update TABLE::lock_type accordingly. */
if (!tl->placeholder() && !using_lock_tables) if (!tl->placeholder() && !using_lock_tables)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7