Problem: MYSQL_BIN_LOG::reset_logs acquires mutexes in wrong order.
The correct order is first LOCK_thread_count and then LOCK_log. This function
does it the other way around. This leads to deadlock when run in parallel
with a thread that takes the two locks in correct order. For example, a thread
that disconnects will take the locks in the correct order.
Fix: change order of the locks in MYSQL_BIN_LOG::reset_logs:
first LOCK_thread_count and then LOCK_log.
This assert could be triggered during two phase commit if binary
log was used as transaction coordinator log. The triggered assert
checks that the same number of transaction IDs are processed in
the prepare and commit phases.
The reason it was triggered, was that the transaction consisted
of an INSERT/UPDATE IGNORE that had an ignorable error. Since it
had an error, no row log events were made and therefore
prepared_xids was 0. However, since it was an IGNORE statement,
the statement started a read/write statement transaction, committed
it and completed successfully.
This patch fixes the problem by adjusting the assert to take
this possibility into account.
Test case added to binlog.binlog_innodb_row.test.
Before this fix, all the performance schema instrumentation for both the binary log
and the relay log would use the following instruments:
- wait/io/file/sql/binlog
- wait/io/file/sql/binlog_index
- wait/synch/mutex/sql/MYSQL_BIN_LOG::LOCK_index
- wait/synch/cond/sql/MYSQL_BIN_LOG::update_cond
This instrumentation is too general and can be more specific.
With this fix, the binlog instrumentation is identical,
and the relay log instrumentation is changed to:
- wait/io/file/sql/relaylog
- wait/io/file/sql/relaylog_index
- wait/synch/mutex/sql/MYSQL_RELAY_LOG::LOCK_index
- wait/synch/cond/sql/MYSQL_RELAY_LOG::update_cond
With this change, the performance instrumentation for the binary log and the relay log,
which share the same structure but have different uses, is more detailed.
This is especially important for hosts in the middle of a replication chain,
that are both masters (binlog) and slaves (relaylog).
From a user perspective, the problem is that a FLUSH LOGS or SIGHUP
signal could end up associating the stdout and stderr to random
files. In the case of this bug report, the streams would end up
associated to InnoDB ibd files.
The freopen(3) function is not thread-safe on FreeBSD. What this
means is that if another thread calls open(2) during freopen()
is executing that another thread's fd returned by open(2) may get
re-associated with the file being passed to freopen(3). See FreeBSD
PR number 79887 for reference:
http://www.freebsd.org/cgi/query-pr.cgi?pr=79887
This problem is worked around by substituting a internal hook within
the FILE structure. This avoids the loss of atomicity by not having
the original fd closed before its duplicated.
Patch based on the original work by Vasil Dimov.
Post-push fixes:
- fixed platform dependent result files
- appeasing valgrind warnings:
Fault injection was also uncovering a previously existing
potential mem leaks. For BUG#46166 testing purposes, fixed
by forcing handling the leak when injecting faults.
Manual merge from mysql-5.1-bugteam into mysql-5.5-bugteam.
Conflicts
=========
Text conflict in sql/log.cc
Text conflict in sql/log.h
Text conflict in sql/slave.cc
Text conflict in sql/sql_parse.cc
Text conflict in sql/sql_priv.h
when generating new name.
If find_uniq_filename returns an error, then this error is not
being propagated upwards, and execution does not report error to
the user (although a entry in the error log is generated).
Additionally, some more errors were ignored in new_file_impl:
- when writing the rotate event
- when reopening the index and binary log file
This patch addresses this by propagating the error up in the
execution stack. Furthermore, when rotation of the binary log
fails, an incident event is written, because there may be a
chance that some changes for a given statement, were not properly
logged. For example, in SBR, LOAD DATA INFILE statement requires
more than one event to be logged, should rotation fail while
logging part of the LOAD DATA events, then the logged data would
become inconsistent with the data in the storage engine.
the DROP statement ..."
Problem: When using temporary tables and closing a session, an
implicit DROP TEMPORARY TABLE IF EXISTS is written to the binary
log (while cleaning up the context of the session THD - see:
sql_class.cc:THD::cleanup which calls close_temporary_tables).
close_temporary_tables, first checks if the binary log is opened
and then proceeds to creating the DROP statements. Then, such
statements, are written to the binary log through
MYSQL_BIN_LOG::write(Log_event *). Inside, there is another check
if the binary log is opened and if not an error is returned. This
is where the faulty behavior is triggered. Given that the test
case replays a binary log, with temp tables statements, and right
after it issues RESET MASTER, there is a chance that is_open will
report false (when the mysql session is closed and the temporary
tables are written).
is_open may return false, because MYSQL_BIN_LOG::reset_logs is
not setting the correct flag (LOG_CLOSE_TO_BE_OPENED), on the
MYSQL_LOG_BIN::log_state (instead it sets just the
LOG_CLOSE_INDEX flag, leaving the log_state to
LOG_CLOSED). Thence, when writing the DROP statement as part of
the THD::cleanup, the thread could get a return value of false
for is_open - inside MYSQL_BIN_LOG::write, ultimately reporting
that it can't write the event to the binary log.
Fix: We fix this by adding the correct flag, missing in the
second close.
Before this fix, file io for the binary log file was not accounted properly,
and showed no io at all.
This bug was due to the following issues:
1) file io for the binlog was instrumented:
- sometime as "wait/io/file/sql/binlog"
- sometime as "wait/io/file/sql/MYSQL_LOG"
leading to inconsistent event_names.
2) the binlog file itself was using an IO_CACHE,
but the IO_CACHE implementation in mysys/mf_iocache.c was
not instrumented to make performance schema calls to record file io.
3) The "wait/io/file/sql/MYSQL_LOG" instrumentation was used
for several log files, such as:
- the binary log
- the slow log
- the query log
which caused file io in these different log files to be accounted
against the same instrument.
The instrumentation needs to have a finer grain and report io
in different event_names, because each file really serves a different purpose.
With this fix:
- the IO_CACHE implementation is now instrumented
- the "wait/io/file/sql/MYSQL_LOG" instrument has been removed
- binlog io is now always instrumented with "wait/io/file/sql/binlog"
- the slow log is instrumented with a new name, "wait/io/file/sql/slow_log"
- the query log is instrumented with a new name, "wait/io/file/sql/query_log"
After the WL#2687, the binlog_cache_size and max_binlog_cache_size affect both the
stmt-cache and the trx-cache. This means that the resource used is twice the amount
expected/defined by the user.
The binlog_cache_use is incremented when the stmt-cache or the trx-cache is used
and binlog_cache_disk_use is incremented when the disk space from the stmt-cache or the
trx-cache is used. This behavior does not allow to distinguish which cache may be harming
performance due to the extra disk accesses and needs to have its in-memory cache
increased.
To fix the problem, we introduced two new options and status variables related to the
stmt-cache:
Options:
. binlog_stmt_cache_size
. max_binlog_stmt_cache_size
Status Variables:
. binlog_stmt_cache_use
. binlog_stmt_cache_disk_use
So there are
. binlog_cache_size that defines the size of the transactional cache for
updates to transactional engines for the binary log.
. binlog_stmt_cache_size that defines the size of the statement cache for
updates to non-transactional engines for the binary log.
. max_binlog_cache_size that sets the total size of the transactional
cache.
. max_binlog_stmt_cache_size that sets the total size of the statement
cache.
. binlog_cache_use that identifies the number of transactions that used the
temporary transactional binary log cache.
. binlog_cache_disk_use that identifies the number of transactions that used
the temporary transactional binary log cache but that exceeded the value of
binlog_cache_size.
. binlog_stmt_cache_use that identifies the number of statements that used the
temporary non-transactional binary log cache.
. binlog_stmt_cache_disk_use that identifies the number of statements that used
the temporary non-transactional binary log cache but that exceeded the value of
binlog_stmt_cache_size.
The binlog_cache_use is incremented twice when changes to a transactional table
are committed, i.e. TC_LOG_BINLOG::log_xid calls is called. The problem happens
because log_xid calls both binlog_flush_stmt_cache and binlog_flush_trx_cache
without checking if such caches are empty thus unintentionally increasing the
binlog_cache_use value twice.
Although not explicitly mentioned in the bug, the binlog_cache_disk_use presents
the same problem.
The binlog_cache_use and binlog_cache_disk_use are status variables that are
incremented when transactional (i.e. trx-cache) or non-transactional (i.e.
stmt-cache) changes are committed. They are used to compute the ratio between
the use of disk and memory while gathering updates for a transaction.
The problem reported is fixed by avoiding incrementing the binlog_cache_use
and binlog_cache_disk_use if a cache is empty. We also have decided to increment
both variables when a cache is truncated as the cache is used although its
content is discarded and is not written to the binary log.
In this patch, we take the opportunity to re-organize the code around the
function binlog_flush_trx_cache and binlog_flush_stmt_cache.
Fix assorted warnings that are generated in optimized builds.
Most of it is silencing variables that are set but unused.
This patch also introduces the MY_ASSERT_UNREACHABLE macro
which helps the compiler to deduce that a certain piece of
code is unreachable.
For crash testing: kill the server without generating core file.
include/my_dbug.h
Use kill(getpid(), SIGKILL) which cannot be caught by signal handlers.
All DBUG_XXX macros should be no-ops in optimized mode, do that for DBUG_ABORT as well.
sql/handler.cc
Kill server without generating core.
sql/log.cc
Kill server without generating core.
For crash testing: kill the server without generating core file.
include/my_dbug.h
Use kill(getpid(), SIGKILL) which cannot be caught by signal handlers.
All DBUG_XXX macros should be no-ops in optimized mode, do that for DBUG_ABORT as well.
sql/handler.cc
Kill server without generating core.
sql/log.cc
Kill server without generating core.
The binlog_cache_use is incremented twice when changes to a transactional table
are committed, i.e. TC_LOG_BINLOG::log_xid calls is called. The problem happens
because log_xid calls both binlog_flush_stmt_cache and binlog_flush_trx_cache
without checking if such caches are empty thus unintentionally increasing the
binlog_cache_use value twice.
To fix the problem we avoided incrementing the binlog_cache_use if the cache is
empty. We also decided to increment binlog_cache_use when the cache is truncated
as the cache is used although its content is discarded and is not written to the
binary log.
Note that binlog_cache_use is incremented for both types of cache, transactional
and non-transactional and that the behavior presented in this patch also applies
to the binlog_cache_disk_use.
Finally, we re-organized the code around the functions binlog_flush_trx_cache and
binlog_flush_stmt_cache.
In sql/log.c, member function wait_for_update_bin_log, a condition is entered with
THD::enter_cond but is not exited. This might leave dangling references to the
mutex/condition in the per-thread information area.
To fix the problem, we call exit_cond to properly remove references to the mutex,
LOCK_log.
temp table
This patch introduces two key changes in the replication's behavior.
Firstly, it reverts part of BUG#51894 which puts any update to temporary tables
into the trx-cache. Now, updates to temporary tables are handled according to
the type of their engines as a regular table.
Secondly, an unsafe mixed statement, (i.e. a statement that access transactional
table as well non-transactional or temporary table, and writes to any of them),
are written into the trx-cache in order to minimize errors in the execution when
the statement logging format is in use.
Such changes has a direct impact on which statements are classified as unsafe
statements and thus part of BUG#53259 is reverted.
After BUG#36649, warnings for sub-statements are cleared when a
new sub-statement is started. This is problematic since it suppresses
warnings for unsafe statements in some cases. It is important that we
always give a warning to the client, because the user needs to know
when there is a risk that the slave goes out of sync.
We fixed the problem by generating warning messages for unsafe statements
while returning from a stored procedure, function, trigger or while
executing a top level statement.
We also started checking unsafeness when both performance and log tables are
used. This is necessary after the performance schema which does a distinction
between performance and log tables.
In sql/log.c, member function wait_for_update_bin_log, a condition is entered with
THD::enter_cond but is not exited. This might leave dangling references to the
mutex/condition in the per-thread information area.
To fix the problem, we call exit_cond to properly remove references to the mutex,
LOCK_log.
A CREATE...SELECT that fails is written to the binary log if a non-transactional
statement is updated. If the logging format is ROW, the CREATE statement and the
changes are written to the binary log as distinct events and by consequence the
created table is not rolled back in the slave.
In this patch, we opted to let the slave goes out of sync by not writting to the
binary log the CREATE statement. We do this by simply reseting the binary log's
cache.
This patch also fixes Bug#55452 "SET PASSWORD is
replicated twice in RBR mode".
The goal of this patch is to remove the release of
metadata locks from close_thread_tables().
This is necessary to not mistakenly release
the locks in the course of a multi-step
operation that involves multiple close_thread_tables()
or close_tables_for_reopen().
On the same token, move statement commit outside
close_thread_tables().
Other cleanups:
Cleanup COM_FIELD_LIST.
Don't call close_thread_tables() in COM_SHUTDOWN -- there
are no open tables there that can be closed (we leave
the locked tables mode in THD destructor, and this
close_thread_tables() won't leave it anyway).
Make open_and_lock_tables() and open_and_lock_tables_derived()
call close_thread_tables() upon failure.
Remove the calls to close_thread_tables() that are now
unnecessary.
Simplify the back off condition in Open_table_context.
Streamline metadata lock handling in LOCK TABLES
implementation.
Add asserts to ensure correct life cycle of
statement transaction in a session.
Remove a piece of dead code that has also become redundant
after the fix for Bug 37521.
Fix warnings flagged by the new warning option -Wunused-but-set-variable
that was added to GCC 4.6 and that is enabled by -Wunused and -Wall. The
option causes a warning whenever a local variable is assigned to but is
later unused. It also warns about meaningless pointer dereferences.
Essentially, the problem is that safemalloc is excruciatingly
slow as it checks all allocated blocks for overrun at each
memory management primitive, yielding a almost exponential
slowdown for the memory management functions (malloc, realloc,
free). The overrun check basically consists of verifying some
bytes of a block for certain magic keys, which catches some
simple forms of overrun. Another minor problem is violation
of aliasing rules and that its own internal list of blocks
is prone to corruption.
Another issue with safemalloc is rather the maintenance cost
as the tool has a significant impact on the server code.
Given the magnitude of memory debuggers available nowadays,
especially those that are provided with the platform malloc
implementation, maintenance of a in-house and largely obsolete
memory debugger becomes a burden that is not worth the effort
due to its slowness and lack of support for detecting more
common forms of heap corruption.
Since there are third-party tools that can provide the same
functionality at a lower or comparable performance cost, the
solution is to simply remove safemalloc. Third-party tools
can provide the same functionality at a lower or comparable
performance cost.
The removal of safemalloc also allows a simplification of the
malloc wrappers, removing quite a bit of kludge: redefinition
of my_malloc, my_free and the removal of the unused second
argument of my_free. Since free() always check whether the
supplied pointer is null, redudant checks are also removed.
Also, this patch adds unit testing for my_malloc and moves
my_realloc implementation into the same file as the other
memory allocation primitives.
