bug #26842: master binary log contains invalid queries - replication fails
bug #12826: Possible to get inconsistent slave using SQL syntax Prepared Statements
Problem:
binlogging PS' we may produce syntacticly incorrect queries in the binlog replacing
some parameters with variable names (instead of variable values).
E.g. in the reported case of "limit ?" clause: replacing "?" with "@var"
produces "limit @var" which is not a correct SQL syntax.
Also it may lead to different query execution on slave if we
set and use a variable in the same statement, e.g.
"insert into t1 values (@x:=@x+1, ?)"
Fix: make the stored statement string created upon its execution use variable values
(instead of names) to fill placeholders.
- The "mysql client in mysqld"(which is used by
replication and federated) should use alarms instead of setting
socket timeout value if the rest of the server uses alarm. By
always calling 'my_net_set_write_timeout'
or 'my_net_set_read_timeout' when changing the timeout value(s), the
selection whether to use alarms or timeouts will be handled by
ifdef's in those two functions.
- Move declaration of 'vio_timeout' into "vio_priv.h"
Problem: we may get syntactically incorrect queries in the binary log
if we use a string value user variable executing a PS which
contains '... limit ?' clause, e.g.
prepare s from "select 1 limit ?";
set @a='qwe'; execute s using @a;
Fix: raise an error in such cases.
is involved.
The Arg_comparator::compare_datetime() comparator caches its arguments if
they are constants i.e. const_item() returns true. The
Item_func_get_user_var::const_item() returns true or false based on
the current query_id and the query_id where the variable was created.
Thus even if a query can change its value its const_item() still will return
true. All this leads to a wrong comparison result when an object of the
Item_func_get_user_var class is involved.
Now the Arg_comparator::can_compare_as_dates() and the
get_datetime_value() functions never cache result of the GET_USER_VAR()
function (the Item_func_get_user_var class).
Made year 2000 handling more uniform
Removed year 2000 handling out from calc_days()
The above removes some bugs in date/datetimes with year between 0 and 200
Now we get a note when we insert a datetime value into a date column
For default values to CREATE, don't give errors for warning level NOTE
Fixed some compiler failures
Added library ws2_32 for windows compilation (needed if we want to compile with IOCP support)
Removed duplicate typedef TIME and replaced it with MYSQL_TIME
Better (more complete) fix for: Bug#21103 "DATE column not compared as DATE"
Fixed properly Bug#18997 "DATE_ADD and DATE_SUB perform year2K autoconversion magic on 4-digit year value"
Fixed Bug#23093 "Implicit conversion of 9912101 to date does not match cast(9912101 as date)"
- Since isinf() portability across various platforms and
compilers is a complicated question, we should not use
it directly. Instead, the my_isinf() macro should be used,
which is defined as an alias to the system-defined isinf()
if it is safe to use, or a workaround implementation otherwise
Bug#21483 "Server abort or deadlock on INSERT DELAYED with another
implicit insert"
Also fixes and adds test cases for bugs:
20497 "Trigger with INSERT DELAYED causes Error 1165"
21714 "Wrong NEW.value and server abort on INSERT DELAYED to a
table with a trigger".
Post-review fixes.
Problem:
In MySQL INSERT DELAYED is a way to pipe all inserts into a
given table through a dedicated thread. This is necessary for
simplistic storage engines like MyISAM, which do not have internal
concurrency control or threading and thus can not
achieve efficient INSERT throughput without support from SQL layer.
DELAYED INSERT works as follows:
For every distinct table, which can accept DELAYED inserts and has
pending data to insert, a dedicated thread is created to write data
to disk. All user connection threads that attempt to
delayed-insert into this table interact with the dedicated thread in
producer/consumer fashion: all records to-be inserted are pushed
into a queue of the dedicated thread, which fetches the records and
writes them.
In this design, client connection threads never open or lock
the delayed insert table.
This functionality was introduced in version 3.23 and does not take
into account existence of triggers, views, or pre-locking.
E.g. if INSERT DELAYED is called from a stored function, which,
in turn, is called from another stored function that uses the delayed
table, a deadlock can occur, because delayed locking by-passes
pre-locking. Besides:
* the delayed thread works directly with the subject table through
the storage engine API and does not invoke triggers
* even if it was patched to invoke triggers, if triggers,
in turn, used other tables, the delayed thread would
have to open and lock involved tables (use pre-locking).
* even if it was patched to use pre-locking, without deadlock
detection the delayed thread could easily lock out user
connection threads in case when the same table is used both
in a trigger and on the right side of the insert query:
the delayed thread would not release locks until all inserts
are complete, and user connection can not complete inserts
without having locks on the tables used on the right side of the
query.
Solution:
These considerations suggest two general alternatives for the
future of INSERT DELAYED:
* it is considered a full-fledged alternative to normal INSERT
* it is regarded as an optimisation that is only relevant
for simplistic engines.
Since we missed our chance to provide complete support of new
features when 5.0 was in development, the first alternative
currently renders infeasible.
However, even the second alternative, which is to detect
new features and convert DELAYED insert into a normal insert,
is not easy to implement.
The catch-22 is that we don't know if the subject table has triggers
or is a view before we open it, and we only open it in the
delayed thread. We don't know if the query involves pre-locking
until we have opened all tables, and we always first create
the delayed thread, and only then open the remaining tables.
This patch detects the problematic scenarios and converts
DELAYED INSERT to a normal INSERT using the following approach:
* if the statement is executed under pre-locking (e.g. from
within a stored function or trigger) or the right
side may require pre-locking, we detect the situation
before creating a delayed insert thread and convert the statement
to a conventional INSERT.
* if the subject table is a view or has triggers, we shutdown
the delayed thread and convert the statement to a conventional
INSERT.
function.
A wrong condition was used to check that the
Arg_comparator::can_compare_as_dates() function calculated the value of the
string constant. When comparing a non-const STRING function with a constant
DATETIME function it leads to saving an arbitrary value as a cached value of
the DATETIME function.
Now the Arg_comparator::set_cmp_func() function initializes the const_value
variable to the impossible DATETIME value (-1) and this const_value is
cached only if it was changed by the Arg_comparator::can_compare_as_dates()
function.
to NULL
For queries of the form SELECT MIN(key_part_k) FROM t1
WHERE key_part_1 = const and ... and key_part_k-1 = const,
the opt_sum_query optimization tries to
use an index to substitute MIN/MAX functions with their values according
to the following rules:
1) Insert the minimum non-null values where the WHERE clause still matches, or
3) A row of nulls
However, the correct semantics requires that there is a third case 2)
such that a NULL value is substituted if there are only NULL values for
key_part_k.
The patch modifies opt_sum_query() to handle this missing case.
for a query over an empty table right after its creation.
The crash is the result of an attempt made by JOIN::optimize to evaluate
the WHERE condition when no records have been actually read.
The added test case can reproduce the crash only with InnoDB tables and
only with 5.0.x.
statement from a UNION query with ORDER BY an expression containing
RAND().
The crash happened because the global order by list in the union query
was not re-initialized for execution.
(Local order by lists were re-initialized though).
a crash when the left operand of the predicate is evaluated to NULL.
It happens when the rows from the inner tables (tables from the subquery)
are accessed by index methods with key values obtained by evaluation of
the left operand of the subquery predicate. When this predicate is
evaluated to NULL an alternative access with full table scan is used
to check whether the result set returned by the subquery is empty or not.
The crash was due to the fact the info about the access methods used for
regular key values was not properly restored after a switch back from the
full scan access method had occurred.
The patch restores this info properly.
The same problem existed for queries with IN subquery predicates if they
were used not at the top level of the queries.
database.
If a user has a right to update anything in the current database then the
access was granted and further checks of access rights for underlying tables
wasn't done correctly. The check is done before a view is opened and thus no
check of access rights for underlying tables can be carried out.
This allows a user to update through a view a table from another database for
which he hasn't enough rights.
Now the mysql_update() and the mysql_test_update() functions are forces
re-checking of access rights after a view is opened.
Bug #20662 "Infinite loop in CREATE TABLE IF NOT EXISTS ... SELECT
with locked tables"
Bug #20903 "Crash when using CREATE TABLE .. SELECT and triggers"
Bug #24738 "CREATE TABLE ... SELECT is not isolated properly"
Bug #24508 "Inconsistent results of CREATE TABLE ... SELECT when
temporary table exists"
Deadlock occured when one tried to execute CREATE TABLE IF NOT
EXISTS ... SELECT statement under LOCK TABLES which held
read lock on target table.
Attempt to execute the same statement for already existing
target table with triggers caused server crashes.
Also concurrent execution of CREATE TABLE ... SELECT statement
and other statements involving target table suffered from
various races (some of which might've led to deadlocks).
Finally, attempt to execute CREATE TABLE ... SELECT in case
when a temporary table with same name was already present
led to the insertion of data into this temporary table and
creation of empty non-temporary table.
All above problems stemmed from the old implementation of CREATE
TABLE ... SELECT in which we created, opened and locked target
table without any special protection in a separate step and not
with the rest of tables used by this statement.
This underminded deadlock-avoidance approach used in server
and created window for races. It also excluded target table
from prelocking causing problems with trigger execution.
The patch solves these problems by implementing new approach to
handling of CREATE TABLE ... SELECT for base tables.
We try to open and lock table to be created at the same time as
the rest of tables used by this statement. If such table does not
exist at this moment we create and place in the table cache special
placeholder for it which prevents its creation or any other usage
by other threads.
We still use old approach for creation of temporary tables.
Also note that we decided to postpone introduction of some tests
for concurrent behaviour of CREATE TABLE ... SELECT till 5.1.
The main reason for this is absence in 5.0 ability to set @@debug
variable at runtime, which can be circumvented only by using several
test files with individual .opt files. Since the latter is likely
to slowdown test-suite unnecessary we chose not to push this tests
into 5.0, but run them manually for this version and later push
their optimized version into 5.1
