Dropping an user defined function may cause server crash in
case this function is still in use by another thread.
The problem was that our hash implementation didn't update
hash link list properly when hash_update() was called.
mysys/hash.c:
The following requirement wasn't met by hash_update() function
causing corruption of hash links list:
After a record was unlinked from the old chain during update, it
holds random position. By the chance this position is equal to
position for the first element in the new chain. That means
updated record is the only record in the new chain.
into olga.mysql.com:/home/igor/mysql-5.0-opt
mysql-test/r/type_enum.result:
Auto merged
mysql-test/t/type_enum.test:
Auto merged
sql/field_conv.cc:
SCCS merged
When a UNION statement forced conversion of an UTF8
charset value to a binary charset value, the byte
length of the result values was truncated to the
CHAR_LENGTH of the original UTF8 value.
sql/item.cc:
Fixed bug #29205.
The calculation of data length was modified in
the Item_type_holder::join_types method to take into
account possible conversion of a multibyte charset
value to a binary charset value, when each
multibyte character is converted into a sequence
of bytes (not to a single byte of binary charset).
mysql-test/t/ctype_utf8.test:
Updated test case for bug #29205.
mysql-test/r/ctype_utf8.result:
Updated test case for bug #29205.
spaces.
When the my_strnncollsp_simple function compares two strings and one is a prefix
of another then this function compares characters in the rest of longer key
with the space character to find whether the longer key is greater or less.
But the sort order of the collation isn't used in this comparison. This may
lead to a wrong comparison result, wrongly created index or wrong order of the
result set of a query with the ORDER BY clause.
Now the my_strnncollsp_simple function uses collation sort order to compare
the characters in the rest of longer key with the space character.
mysql-test/t/ctype_collate.test:
Added a test case for the bug#29261: Sort order of the collation wasn't used
when comparing trailing spaces.
mysql-test/r/ctype_collate.result:
Added a test case for the bug#29261: Sort order of the collation wasn't used
when comparing trailing spaces.
strings/ctype-simple.c:
Bug#29261: Sort order of the collation wasn't used when comparing trailing
spaces.
Now the my_strnncollsp_simple function uses collation sort order to compare
the characters in the rest of longer key with the space character.
mysql-test/r/ps_1general.result:
Update result file.
mysql-test/r/ps_2myisam.result:
Update result file.
mysql-test/r/ps_3innodb.result:
Update result file.
mysql-test/r/ps_4heap.result:
Update result file.
mysql-test/r/ps_5merge.result:
Update result file.
tests/mysql_client_test.c:
Fix test -- after field changing character set to utf8 in the server,
length should be calculated differently.
query / no aggregate of subquery
The optimizer counts the aggregate functions that
appear as top level expressions (in all_fields) in
the current subquery. Later it makes a list of these
that it uses to actually execute the aggregates in
end_send_group().
That count is used in several places as a flag whether
there are aggregates functions.
While collecting the above info it must not consider
aggregates that are not aggregated in the current
context. It must treat them as normal expressions
instead. Not doing that leads to incorrect data about
the query, e.g. running a query that actually has no
aggregate functions as if it has some (and hence is
expected to return only one row).
Fixed by ignoring the aggregates that are not aggregated
in the current context.
One other smaller omission discovered and fixed in the
process : the place of aggregation was not calculated for
user defined functions. Fixed by calling
Item_sum::init_sum_func_check() and
Item_sum::check_sum_func() as it's done for the rest of
the aggregate functions.
mysql-test/r/subselect.result:
Bug #27333: test case
mysql-test/t/subselect.test:
Bug #27333: test case
sql/item_subselect.cc:
Bug#27333: need select_lex to filter out
aggregates that are not aggregated in
the current select.
sql/item_sum.cc:
Bug#27333: need select_lex to filter out
aggregates that are not aggregated in
the current select.
sql/item_sum.h:
Bug#27333: calculate the place of
aggregation for user defined functions.
sql/sql_select.cc:
Bug#27333: When counting the aggregated functions
and collecting a list of them we must not consider
the aggregates that are not aggregated in the local
context as "local" : i.e. we must treat them as
normal functions and not add them to the aggregate
functions list.
sql/sql_select.h:
Bug#27333: need select_lex to filter out
aggregates that are not aggregated in
the current select.
"Federared Transactions Failure"
Bug occurs when the user performs an operation which inserts more than
one row into the federated table and the federated table references a
remote table stored within a transactional storage engine. When the
insert operation for any one row in the statement fails due to
constraint violation, the federated engine is unable to perform
statement rollback and so the remote table contains a partial commit.
The user would expect a statement to perform the same so a statement
rollback is expected.
This bug was fixed by implementing bulk-insert handling into the
federated storage engine. This will relieve the bug for most common
situations by enabling the generation of a multi-row insert into the
remote table and thus permitting the remote table to perform
statement rollback when neccessary.
The multi-row insert is limited to the maximum packet size between
servers and should the size overflow, more than one insert statement
will be sent and this bug will reappear. Multi-row insert is disabled
when an "INSERT...ON DUPLICATE KEY UPDATE" is being performed.
The bulk-insert handling will offer a significant performance boost
when inserting a large number of small rows.
This patch builds on Bug29019 and Bug25511
sql/ha_federated.cc:
bug25513
new member methods:
start_bulk_insert() - initializes memory for bulk insert
end_bulk_insert() - sends any remaining bulk insert and frees memory
append_stmt_insert() - create the INSERT statement
sql/ha_federated.h:
bug25513
new member value:
bulk_insert
new member methods:
start_bulk_insert(), end_bulk_insert(), append_stmt_insert()
make member methods private:
read_next(), index_read_idx_with_result_set()
mysql-test/r/federated_innodb.result:
New BitKeeper file ``mysql-test/r/federated_innodb.result''
mysql-test/t/federated_innodb-slave.opt:
New BitKeeper file ``mysql-test/t/federated_innodb-slave.opt''
mysql-test/t/federated_innodb.test:
New BitKeeper file ``mysql-test/t/federated_innodb.test''
"Federated INSERT failures"
Federated does not correctly handle "INSERT...ON DUPLICATE KEY UPDATE"
However, implementing such support is not reasonably possible without
increasing complexity of the storage engine: checking that constraints
on remote server match local server and parsing error messages.
This patch causes 'ON DUPLICATE KEY' to fail with ER_DUP_KEY message
if a conflict occurs and not to fail silently.
include/my_base.h:
bug25511
new storage engine hint: HA_EXTRA_INSERT_WITH_UPDATE
mysql-test/r/federated.result:
test for bug25511
mysql-test/t/federated.test:
test for bug25511
sql/ha_federated.cc:
bug25511
implement support for handling HA_EXTRA_INSERT_WITH_UPDATE hint
sql/ha_federated.h:
bug25511
new property: insert_dup_update
sql/sql_insert.cc:
bug25511
implement support for HA_EXTRA_INSERT_WITH_UPDATE
When checking duplicates flag, if it is DUP_UPDATE, send hint
to the storage engine.
SHOW CREATE TABLE or SELECT FROM I_S.
Actually, the bug discovers two problems:
- the original query is not preserved properly. This is the problem
of BUG#16291;
- the resultset of SHOW CREATE TABLE statement is binary.
This patch fixes the second problem for the 5.0.
Both problems will be fixed in 5.1.
mysql-test/r/show_check.result:
Update result file.
mysql-test/t/show_check.test:
Provide test case for BUG#10491.
sql/item.h:
Use utf8_general_ci instead of binary collation by default,
because for views and base tables utf8 is the character set
in which their definition is stored. For system constants
it's the default character set, and for other objects
(routines, triggers), no character set is stored, and
therefore no character set is known, so returning utf8
is just as good as any non-binary character set.
This latter problem is fixed in 5.1 by 16291. In 5.1
we will return the "real" character set.
Problem: like_range() returned wrong ranges for contractions (like 'ch' in Czech').
Fix: adding a special code to handle tricky cases:
- contraction head followed by a wild character
- full contraction
- contraction part followed by another contraction part,
but they are not a contraction together.
mysql-test/r/ctype_uca.result:
Adding test case
mysql-test/t/ctype_uca.test:
Adding test case
strings/ctype-mb.c:
Adding test case
strings/ctype-uca.c:
Allocate additional 256 bytes for flags "is contraction part".
strings/ctype-ucs2.c:
Adding test case
"REPLACE/INSERT IGNORE/UPDATE IGNORE doesn't work"
Federated does not record neccessary HA_EXTRA flags in order to
support REPLACE/INSERT IGNORE/UPDATE IGNORE.
Implement ::extra() to capture flags neccessary for functionality.
New function append_ident() to better escape identifiers consistantly.
mysql-test/r/federated.result:
test for bug29019
mysql-test/t/federated.test:
test for bug29019
sql/ha_federated.cc:
Bug29019
Federated does not record neccessary HA_EXTRA flags in order to
support REPLACE/INSERT IGNORE/UPDATE IGNORE.
Implement ::extra() to capture flags neccessary for functionality.
New function append_ident() to better escape identifiers consistantly.
sql/ha_federated.h:
bug29019
add 2 member values to ha_federated class
ignore_duplicates and replace_duplicates.
add 1 member method to ha_federated class
extra()
Fulltext index may get corrupt by certain gbk characters.
The problem was that when skipping leading non-true-word-characters,
we assumed that these characters are always 1 byte long. This is not
the case with gbk character set, since non-true-word-characters may
be 2 bytes long.
Affects 5.0 only.
myisam/ft_parser.c:
Leading non-true-word-characters may also be multi-byte (e.g. in
gbk character set).
mysql-test/r/fulltext2.result:
A test case for BUG#29299.
mysql-test/t/fulltext2.test:
A test case for BUG#29299.
Thanks to Martin Friebe for finding and submitting a fix for this bug!
A table with maximum number of key segments and maximum length key name
would have a corrupted .frm file, due to an incorrect calculation of the
complete key length. Now the key length is computed correctly (I hope) :-)
MyISAM would reject a table with the maximum number of keys and the maximum
number of key segments in all keys. It would allow one less than this total
maximum. Now MyISAM accepts a table defined with the maximum. (This is a
very minor issue.)
myisam/mi_open.c:
Bug #26642: change >= to > in a comparison (i.e., error
only if key_parts_in_table really is greater than
MAX_KEY * MAX_KEY_SEG)
mysql-test/r/create.result:
Bug #26642: test case
mysql-test/t/create.test:
Bug #26642: test case
sql/table.cc:
Bug #26642: In create_frm(), fix formula for key_length;
it was too small by (keys * 2) bytes
CHECK TABLE against ARCHIVE table may falsely report table corruption,
or cause server crash.
Fixed by using proper buffer for CHECK TABLE.
Affects both 5.0 and 5.1.
mysql-test/r/archive.result:
A test case for BUG#28916.
mysql-test/t/archive.test:
A test case for BUG#28916.
sql/ha_archive.cc:
We call Field::get_length() from get_row(). Field::get_length() assumes
that the row was read into table->record[0] buffer, which is not the
case when we check a table. As a result we get wrongly initialized
blob length.
Use table->record[0] as record buffer for check table instead.
Sometimes special 0 ENUM values was ALTERed to normal
empty string ENUM values.
Special 0 ENUM value has the same string representation
as normal ENUM value defined as '' (empty string).
The do_field_string function was used to convert
ENUM data at an ALTER TABLE request, but this
function doesn't care about numerical "indices" of
ENUM values, i.e. do_field_string doesn't distinguish
a special 0 value from an empty string value.
A new copy function called do_field_enum has been added to
copy special 0 ENUM values without conversion to an empty
string.
sql/field_conv.cc:
Fixed bug #29251.
The Copy_field::get_copy_func method has been modified to
return a pointer to the do_field_enum function if a conversion
between two columns of incompatible enum types is required.
The do_field_enum function has been added for the correct
conversion of special 0 enum values.
mysql-test/t/type_enum.test:
Updated test case for bug #29251.
mysql-test/r/type_enum.result:
Updated test case for bug #29251.
a lookup into a BINARY index by a key ended with spaces. It caused
an assertion abort for a debug version and wrong results for non-debug
versions.
The problem occurred because the function _mi_pack_key stripped off
the trailing spaces from binary search keys while the function _mi_make_key
did not do it when keys were inserted into the index.
Now the function _mi_pack_key does not remove the trailing spaces from
search keys if they are of the binary type.
mysql-test/r/binary.result:
Added a test case for bug #29087.
mysql-test/t/binary.test:
Added a test case for bug #29087.
If one sets MYSQL_READ_DEFAULTS_FILE and MYSQL_READ_DEFAULT_GROUP options
after mysql_real_connect() called with that MYSQL instance,
these options will affect next mysql_reconnect then.
As we use a copy of the original MYSQL object inside mysql_reconnect,
and mysql_real_connect frees options.my_cnf_file and _group strings,
we will free these twice when we execute mysql_reconnect with the
same MYSQL for the second time.
I don't think we should ever read defaults files handling mysql_reconnect.
So i just set them to 0 for the temporary MYSQL object there/
sql-common/client.c:
Bug #29247 Double free in libmysqlclient_r when mysql restarted.
we don't need mysql_real_connect to reread defaults file in this
case, so set related parameters to zero
