erroneous check
Problem: Actually there were two problems in the server code. The check
for SQLCOM_FLUSH in SF/Triggers were not according to the existing
architecture which uses sp_get_flags_for_command() from sp_head.cc .
This function was also missing a check for SQLCOM_FLUSH which has a
problem combined with prelocking. This changeset fixes both of these
deficiencies as well as the erroneous check in
sp_head::is_not_allowed_in_function() which was a copy&paste error.
mysql-test/r/sp-error.result:
update result
mysql-test/r/trigger.result:
update result
mysql-test/t/sp-error.test:
FLUSH can create a problem with prelocking, hence it's disabled.
There is a better way to check this than a check in the parser.
Now we use sp_get_flags_for_command() and the error returned is
different.
mysql-test/t/trigger.test:
FLUSH can create a problem with prelocking, hence it's disabled.
There is a better way to check this than a check in the parser.
Now we use sp_get_flags_for_command() and the error returned is
different.
sql/sp_head.cc:
FLUSH and RESET are not allowed inside a SF/Trigger.
Because they don't imply a COMMIT sp_head::HAS_COMMIT_OR_ROLLBACK
cannot be used. Two new flags were introduced for that reason.
sql/sp_head.h:
Don't check m_type as this check is erroneous. This is probably
a copy and paste error when moving code from somewhere else. Another
fact which supports this was prefixing the enum value with the name
of class sp_head.
Adding two new flags HAS_SQLCOM_RESET and HAS_SQLCOM_FLUSH. The values
are 2048 and 4096 because in the 5.1 branch there are already new flags
which are with values up-to 1024.
sql/sql_parse.cc:
FLUSH can cause a problem with prelocking in SF/Trigger and
therefore is already disabled. RESET is also disabled because
is handled by the same code as FLUSH. We won't allow RESET inside
SF/Trigger at that stage without thorough analysis. The check for
them is already done in the parser by calling
is_not_allowed_in_function()
sql/sql_yacc.yy:
By listing SQLCOM_FLUSH as command which implies COMMIT
in sp_get_flags_for_command() the check in sql_yacc.yy is
obsolete.
const tables. This resulted in choosing extremely inefficient
execution plans in same cases when distribution of data in
joined were skewed (see the customer test case for the bug).
mysql-test/r/select.result:
Added a test case for bug #21390: wrong estimate of rows
after elimination of const tables.
Includded a test case that checks the code added by the patch
that handles outer joins with no matches after substitution of
a const table in an efficient way.
mysql-test/t/select.test:
Added a test case for bug #21390: wrong estimate of rows
after elimination of const tables.
Included a test case that checks the code added by the patch
that handles outer joins with no matches after substitution of
a const table in an efficient way.
sql/sql_select.cc:
Fixed bug #21390: wrong estimate of rows after elimination of
const tables. This resulted in choosing extremely inefficient
execution plans in same cases when distribution of data in
joined were skewed (see the customer test case for the bug).
Also added the code to handle outer joins with no matches after
substitution of a const table in an efficient way.
Corrected calculation of the null rejecting key conditions.
The following procedure was not possible if max_sp_recursion_depth is 0
create procedure show_proc() show create procedure show_proc;
Actually there is no recursive call but the limit is checked.
Solved by temporarily increasing the thread's limit just before the fetch from cache
and decreasing after that.
mysql-test/r/sp.result:
update result
mysql-test/t/sp.test:
Test for bug #21416 SP: Recursion level higher than zero needed for non-recursive call
sql/sp.cc:
Increase the max_sp_recursion_depth temporarily for SHOW CREATE PROCEDURE call.
This call is in fact not recursive but is counted as such. Outcome, it will work
always but if max_sp_recursion_depth is reached we are going to cache one more
sp_head instance.
The mysql client uses the default character set on reconnect. The default character set is now controled by the client charset command while the client is running. The charset command now also issues a SET NAMES command to the server to make sure that the client's charset settings are in sync with the server's.
client/mysql.cc:
Client charset command now changes the default character set and issues a SET NAMES command to the server.
mysql-test/r/mysql.result:
Corrected results for new behaviour.
mysql-test/t/mysql.test:
Removed redundant commands from test.
Must not use Item_direct_ref in HAVING because it points to
the new value (witch is not yet calculated for the first row).
mysql-test/r/subselect.result:
Bug #16255: Subquery in where
- test case
mysql-test/t/subselect.test:
Bug #16255: Subquery in where
- test case
sql/item_subselect.cc:
Bug #16255: Subquery in where
Must not use Item_direct_ref in HAVING because it points to
the new value (witch is not yet calculated for the first row).
Changes in an item tree done by optimizer weren't properly
registered and went unnoticed, which resulted in preliminary freeing
of used memory.
mysql-test/r/ps.result:
Add result for bug#21166: Prepared statement causes signal 11
on second execution.
mysql-test/t/ps.test:
Add test case for bug#21166: Prepared statement causes signal 11
on second execution.
sql/item.cc:
Move Item::transform() and Item_default_value::transform() from
item.h here and use THD::change_item_tree() instead of plain
assignment.
Change Item_field::set_no_const_sub() to be used with Item::walk()
instead of Item::transform().
sql/item.h:
Move definition of Item::transform() and Item_default_value::transform()
to item.cc.
Change Item::set_no_const_sub() to be used with Item::walk()
instead of Item::transform().
sql/item_cmpfunc.cc:
Use Item::walk() to execute Item::set_no_const_sub().
Use THD::change_item_tree() instead of plain assignment.
sql/item_func.cc:
Add assert and comment to Item_func::traverse_cond().
sql/item_row.cc:
Use THD::change_item_tree() instead of plain assignment.
sql/item_strfunc.cc:
Move Item_func_make_set::transform() from item_strfunc.h here and use
THD::change_item_tree() instead of plain assignment.
sql/item_strfunc.h:
Move definition of Item_func_make_set::transform() to item_strfunc.cc.
- Fix typo in Item_func_export_set::fix_length_and_dec() which caused character set aggregation to fail
- Remove default argument from last arg of agg_arg_charsets() function, to reduce potential errors
mysql-test/r/func_misc.result:
Test EXPORT_SET() with charset coersion (bug #21531)
mysql-test/t/func_misc.test:
Test EXPORT_SET() with charset coersion (bug #21531)
sql/item_func.h:
Remove default argument from last arg of agg_arg_charsets() function, to reduce potential errors.
sql/item_strfunc.cc:
Fix typo in Item_func_export_set::fix_length_and_dec() which caused character set aggregation to fail.
Corrected test case for the bug#21261
sql_parse.cc:
Corrected fix for bug#21261
mysql-test/t/view.test:
Corrected test case for the bug#21261
mysql-test/r/view.result:
Corrected test case for the bug#21261
sql/sql_parse.cc:
Corrected fix for bug#21261
into siva.hindu.god:/usr/home/tim/m/bk/41
sql/item_strfunc.cc:
Auto merged
sql/item_strfunc.h:
Auto merged
sql/slave.cc:
Auto merged
sql/sql_acl.cc:
Auto merged
User name (host name) has limit on length. The server code relies on these
limits when storing the names. The problem was that sometimes these limits
were not checked properly, so that could lead to buffer overflow.
The fix is to check length of user/host name in parser and if string is too
long, throw an error.
mysql-test/r/grant.result:
Updated result file.
mysql-test/r/sp.result:
Updated result file.
mysql-test/r/trigger.result:
Updated result file.
mysql-test/r/view.result:
Updated result file.
mysql-test/t/grant.test:
Added test for BUG#16899.
mysql-test/t/sp.test:
Added test for BUG#16899.
mysql-test/t/trigger.test:
Added test for BUG#16899.
mysql-test/t/view.test:
Added test for BUG#16899.
sql/mysql_priv.h:
Added prototype for new function.
sql/share/errmsg.txt:
Added new resources.
sql/sql_acl.cc:
Remove outdated checks.
sql/sql_parse.cc:
Add a new function for checking string length.
sql/sql_yacc.yy:
Check length of user/host name.
into lamia.home:/home/timka/mysql/src/5.0-bug-21456
mysql-test/r/distinct.result:
Merge the fix for BUG#21456
mysql-test/t/distinct.test:
Merge the fix for BUG#21456
sql/sql_select.cc:
Merge the fix for BUG#21456
GROUP BY/DISTINCT pruning optimization must be done before ORDER BY
optimization because ORDER BY may be removed when GROUP BY/DISTINCT
sorts as a side effect, e.g. in
SELECT DISTINCT <non-key-col>,<pk> FROM t1
ORDER BY <non-key-col> DISTINCT
must be removed before ORDER BY as if done the other way around
it will remove both.
mysql-test/r/distinct.result:
Test for BUG#21456.
mysql-test/t/distinct.test:
Test for BUG#21456.
sql/sql_select.cc:
Bug #21456: SELECT DISTINCT(x) produces incorrect results when using order by
GROUP BY/DISTINCT pruning optimization must be done before ORDER BY
optimization because ORDER BY may be removed when GROUP BY/DISTINCT
sorts as a side effect.
used.
Sorting by RAND() uses a temporary table in order to get a correct results.
User defined variable was set during filling the temporary table and later
on it is substituted for its value from the temporary table. Due to this
it contains the last value stored in the temporary table.
Now if the result_field is set for the Item_func_set_user_var object it
updates variable from the result_field value when being sent to a client.
The Item_func_set_user_var::check() now accepts a use_result_field
parameter. Depending on its value the result_field or the args[0] is used
to get current value.
mysql-test/r/user_var.result:
Added a test case for bug#16861: User defined variable can have a wrong value if a tmp table was used.
mysql-test/t/user_var.test:
Added a test case for bug#16861: User defined variable can have a wrong value if a tmp table was used.
sql/item_func.cc:
Fixed bug#16861: User defined variable can have a wrong value if a tmp table was used.
Now if the result_field is set for the Item_func_set_user_var object it
updates variable from the result_field value when being sent to a client.
The Item_func_set_user_var::check() now accepts a use_result_field
parameter. Depending on its value the result_field or the args[0] is used
to get current value.
sql/item_func.h:
Fixed bug#16861: User defined variable can have a wrong value if a tmp table was used.
Added a new SUSERVAR_FUNC function type.
Updated the Item_func_set_user_var::check() function declaration.
Added the Item_func_set_user_var::send() member function.
sql/set_var.cc:
Fixed bug#16861: User defined variable can have a wrong value if a tmp table was used.
Modified to use updated Item_func_set_user_var::check() function.
sql/sql_class.cc:
Fixed bug#16861: User defined variable can have a wrong value if a tmp table was used.
Modified to use updated Item_func_set_user_var::check() function.
sql/sql_select.cc:
Fixed bug#16861: User defined variable can have a wrong value if a tmp table was used.
Now an Item_func_set_user_var object isn't substituted for an Item_field object after filling a temporary table.
when X.509 subject was required for a connect, we tested whether it was the right
one, but did not refuse the connexion if not. fixed.
(corrected CS now --replace_results socket-path)
mysql-test/r/openssl_1.result:
Bug#20411: "GRANT ... REQUIRE ISSUER nnn AND SUBJECT mmm" fails to require both
test not only whether we can connect with a correct X.509 subject when one is
required, but also assure that we can't without one.
mysql-test/t/openssl_1.test:
Bug#20411: "GRANT ... REQUIRE ISSUER nnn AND SUBJECT mmm" fails to require both
test not only whether we can connect with a correct X.509 subject when one is
required, but also assure that we can't without one.
sql/sql_acl.cc:
Bug#20411: "GRANT ... REQUIRE ISSUER nnn AND SUBJECT mmm" fails to require both
actually refuse connexion if X.509 is required, but does not match. kudos to Al Smith.
Bug #18744 Test 'join_outer' fails if "classic" configuration in 5.0
- moved an InnoDB dependent test to the appropriate file
mysql-test/r/innodb_mysql.result:
Bug #18744 Test 'join_outer' fails if "classic" configuration in 5.0
- moved an InnoDB dependent test to the appropriate file
mysql-test/r/join_outer.result:
Bug #18744 Test 'join_outer' fails if "classic" configuration in 5.0
- moved an InnoDB dependent test to the appropriate file
mysql-test/t/innodb_mysql.test:
Bug #18744 Test 'join_outer' fails if "classic" configuration in 5.0
- moved an InnoDB dependent test to the appropriate file
mysql-test/t/join_outer.test:
Bug #18744 Test 'join_outer' fails if "classic" configuration in 5.0
- moved an InnoDB dependent test to the appropriate file
server to crash".
Crash caused by assertion failure happened when one ran SHOW OPEN TABLES
while concurrently doing DROP TABLE (or RENAME TABLE, CREATE TABLE LIKE
or any other command that takes name-lock) in other connection.
For non-debug version of server problem exposed itself as wrong output
of SHOW OPEN TABLES statement (it was missing name-locked tables).
Finally in 5.1 both debug and non-debug versions simply crashed in
this situation due to NULL-pointer dereference.
This problem was caused by the fact that table placeholders which were
added to table cache in order to obtain name-lock had TABLE_SHARE::table_name
set to 0. Therefore they broke assumption that this member is non-0 for
all tables in table cache which was checked by assert in list_open_tables()
(in 5.1 this function simply relies on it).
The fix simply sets this member for such placeholders to appropriate value
making this assumption true again.
This patch also includes test for similar bug 12212 "Crash that happens
during removing of database name from cache" reappeared in 5.1 as bug 19403.
mysql-test/r/drop.result:
Added test for bug#21216 "Simultaneous DROP TABLE and SHOW OPEN TABLES
causes server to crash" and bug#12212/19403 "Crash that happens during
removing of database name from cache".
mysql-test/t/drop.test:
Added test for bug#21216 "Simultaneous DROP TABLE and SHOW OPEN TABLES
causes server to crash" and bug#12212/19403 "Crash that happens during
removing of database name from cache".
sql/lock.cc:
lock_table_name():
Our code assumes that TABLE_SHARE::table_name for objects in table cache
is non-NULL (for example look at assertion in list_open_tables()). This
was not true for table placeholders that were added to table cache for
name-locking. So let us set this member for such placeholders.
A date can be represented as an int (like 20060101) and as a string (like
"2006.01.01"). When a DATE/TIME field is compared in one SELECT against both
representations the constant propagation mechanism leads to comparison
of DATE as a string and DATE as an int. In this example it compares 2006 and
20060101 integers. Obviously it fails comparison although they represents the
same date.
Now the Item_bool_func2::fix_length_and_dec() function sets the comparison
context for items being compared. I.e. if items compared as strings the
comparison context is STRING.
The constant propagation mechanism now doesn't mix items used in different
comparison contexts. The context check is done in the
Item_field::equal_fields_propagator() and in the change_cond_ref_to_const()
functions.
Also the better fix for bug 21159 is introduced.
mysql-test/t/type_datetime.test:
Added a test case for bug#21475: Wrongly applied constant propagation leads to a false comparison.
mysql-test/r/type_datetime.result:
Added a test case for bug#21475: Wrongly applied constant propagation leads to a false comparison.
sql/sql_select.cc:
Fixed bug#21475: Wrongly applied constant propagation leads to a false comparison.
The constant propagation mechanism now doesn't mix items used in different
comparison contexts. The check is done in the change_cond_ref_to_const() function.
sql/item_cmpfunc.cc:
Fixed bug#21475: Wrongly applied constant propagation leads to a false comparison.
Now the Item_bool_func2::fix_length_and_dec() function sets the comparison
context for items being compared.
sql/item.h:
Fixed bug#21475: Wrongly applied constant propagation leads to a false comparison.
To the Item class a new field called cmp_context is added.
It represents the comparison context of an item.
sql/item.cc:
Fixed bug#21475: Wrongly applied constant propagation leads to a false comparison.
The constant propagation mechanism now doesn't mix items used in different
comparison contexts. The context check is done in the
Item_field::equal_fields_propagator() function.
The problem was that the error handling was using a too-small buffer to
print the error message generated. We fix this by not using a buffer at
all, but by using fprintf() directly. There were also some problems with
the error handling in table dumping that was exposed by this fix that were
also corrected.
client/mysqldump.c:
Use fprintf() instead of my_printf_error() to avoid buffer overflow issues.
Since ME_BELL wasn't specified, calling my_printf_error() offered no advantage
except for adding my_progname, which we just go ahead and do manually. Also,
fix the error handling in dumpTable() when queries to get data fail and --force
was specified.
mysql-test/r/mysqldump.result:
Add new results
mysql-test/t/mysqldump.test:
Add new regression test
into zippy.cornsilk.net:/home/cmiller/work/mysql/merge/mysql-5.0
BitKeeper/deleted/.del-mysql_client.result:
Auto merged
BitKeeper/deleted/.del-mysql_client.test:
Auto merged
client/mysql.cc:
Auto merged
mysql-test/r/func_time.result:
Auto merged
mysql-test/r/grant.result:
Auto merged
mysql-test/t/func_time.test:
Auto merged
mysql-test/t/grant.test:
Auto merged
sql/item_timefunc.cc:
Auto merged
sql/mysqld.cc:
Auto merged
sql/slave.cc:
Auto merged
sql/sql_acl.cc:
Auto merged
Corrected test case result after fix for bug#18165
view.result, view.test:
Corrected test case for bug#21261
mysql-test/t/view.test:
Corrected test case for bug#21261
mysql-test/r/view.result:
Corrected test case for bug#21261
mysql-test/r/ndb_condition_pushdown.result:
Corrected test case result after fix for bug#18165
into zippy.cornsilk.net:/home/cmiller/work/mysql/merge/mysql-5.0
mysql-test/r/grant.result:
Auto merged
mysql-test/t/grant.test:
Auto merged
tests/mysql_client_test.c:
Auto merged
into zippy.cornsilk.net:/home/cmiller/work/mysql/merge/mysql-5.0
mysql-test/r/grant.result:
Auto merged
mysql-test/t/grant.test:
Auto merged
sql/mysqld.cc:
Auto merged
sql/sql_acl.cc:
Auto merged
mysql-test/r/mysql_client.result:
Removed OS dependant output.
mysql-test/t/mysql_client.test:
Re-directed output to a temporary file because the actual output isn't important and was causing a Windows test failure.
Made [NOT]BETWEEN predicates SARGable in respect to the second and
the third arguments.
mysql-test/r/range.result:
Added a test case to bug #18165.
mysql-test/t/range.test:
Added a test case to bug #18165.
sql/opt_range.cc:
Fixed bug #18165.
Made [NOT]BETWEEN predicates SARGable in respect to the second and
the third arguments.
Put in a separate function called get_full_func_mm_tree the functionality
that builds a conjunction of all SEL_TREEs for a simple predicate of the
form (f op c), where f was a field and c was a constant, applying different
equalities f=f' with f' being another field.
