| Commit message (Collapse) | Author | Age | Files | Lines |
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
A side effect of MDEV-16264 is that a large number of threads will
be created at server startup, to be destroyed after a minute or two.
One source of such thread creation is srv_start_periodic_timer().
InnoDB is creating 3 periodic tasks: srv_master_callback (1Hz)
srv_error_monitor_task (1Hz), and srv_monitor_task (0.2Hz).
It appears that we can merge srv_error_monitor_task and srv_monitor_task
and have them invoked 4 times per minute (every 15 seconds). This will
affect our ability to enforce innodb_fatal_semaphore_wait_threshold and
some computations around BUF_LRU_STAT_N_INTERVAL.
We could remove srv_master_callback along with the DROP TABLE queue
at some point of time in the future. We must keep it independent
of the innodb_fatal_semaphore_wait_threshold detection, because
the background DROP TABLE queue could get stuck due to dict_sys
being locked by another thread. For now, srv_master_callback
must be invoked once per second, so that
innodb_flush_log_at_timeout=1 can work.
BUF_LRU_STAT_N_INTERVAL: Reduce the precision and extend the time
from 50*1 second to 4*15 seconds.
srv_error_monitor_timer: Remove.
MAX_MUTEX_NOWAIT: Increase from 20*1 second to 2*15 seconds.
srv_refresh_innodb_monitor_stats(): Avoid a repeated call to time(NULL).
Change the interval to less than 60 seconds.
srv_monitor(): Renamed from srv_monitor_task.
srv_monitor_task(): Renamed from srv_error_monitor_task().
Invoked only once in 15 seconds. Invoke also srv_monitor().
Increase the fatal_cnt threshold from 10*1 second to 1*15 seconds.
sync_array_print_long_waits_low(): Invoke time(NULL) only once.
Remove a bogus message about printouts for 30 seconds. Those
printouts were effectively already disabled in MDEV-16264
(commit 5e62b6a5e06eb02cbde1e34e95e26f42d87fce02).
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The purpose of the InnoDB page cleaner subsystem is to write out
modified pages from the buffer pool to data files. When the
innodb_max_dirty_pages_pct_lwm is not exceeded or
innodb_adaptive_flushing=ON decides not to write out anything,
the page cleaner should keep sleeping indefinitely until the state
of the system changes: a dirty page is added to the buffer pool such
that the page cleaner would no longer be idle.
buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle.
When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait().
buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed.
innodb_max_dirty_pages_pct_update(),
innodb_max_dirty_pages_pct_lwm_update():
Wake up the page cleaner just in case.
Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are
already waking up the page cleaner thread.
|
| | |
|
| |
| |
| |
| | |
Kudos to Marko for finding.
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
This partially reverts commit 6479006e14691ff85072d06682f81b90875e9cb0.
Remove the constant tpool::aio::N_PENDING, which has no
intrinsic meaning for the tpool.
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
tpool::aio::N_PENDING: Replaces OS_AIO_N_PENDING_IOS_PER_THREAD.
This limits two similar things: the number of outstanding requests
that a thread may io_submit(), and the number of completed requests
collected at a time by io_getevents().
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In the asynchronous I/O interface, InnoDB is invoking io_getevents()
with a timeout value of half a second, and requesting exactly 1 event
at a time.
The reason to have such a short timeout is to facilitate shutdown.
We can do better: Use an infinite timeout, wait for a larger maximum
number of events. On shutdown, we will invoke io_destroy(), which
should lead to the io_getevents system call reporting EINVAL.
my_getevents(): Reimplement the libaio io_getevents() by only invoking
the system call. The library implementation would try to elide the
system call and return 0 immediately if aio_ring_is_empty() holds.
Here, we do want a blocking system call, not 100% CPU usage. Neither
do we want the aio_ring_is_empty() trigger SIGSEGV because it is
dereferencing some memory that was freed by io_destroy().
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
For reading trx_t::state we can avoid acquiring trx_t::mutex.
Atomic load and store should be similar to normal load and store
on most instruction set architectures. The atomicity of the operation
would merely prohibit the compiler from reordering some operations.
|
| | |
|
| |
| |
| |
| |
| |
| | |
We must avoid acquiring a latch while we are already holding one.
The tablespace latch was being acquired recursively in some
operations that allocate or free pages.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
fts_cache_t::init_lock: Replace with mutex. This was only acquired
in exclusive mode.
fts_cache_t::lock: Replace with mutex. The only read-lock user was
i_s_fts_index_cache_fill() for producing content for the view
INFORMATION_SCHEMA.INNODB_FT_INDEX_CACHE.
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Many InnoDB rw-locks unnecessarily depend on the complex
InnoDB rw_lock_t implementation that support the SX lock mode
as well as recursive acquisition of X or SX locks.
One of them is the bunch of adaptive hash index search latches,
instrumented as btr_search_latch in PERFORMANCE_SCHEMA.
Let us introduce a simpler lock for those in order to
reduce overhead.
srw_lock: A simple read-write lock that does not support recursion.
On Microsoft Windows, this wraps SRWLOCK, only adding
runtime overhead if PERFORMANCE_SCHEMA is enabled.
On Linux (all architectures), this is implemented with
std::atomic<uint32_t> and the futex system call.
On other platforms, we will wrap mysql_rwlock_t with
zero runtime overhead.
The PERFORMANCE_SCHEMA instrumentation differs
from InnoDB rw_lock_t in that we will only invoke
PSI_RWLOCK_CALL(start_rwlock_wrwait) or
PSI_RWLOCK_CALL(start_rwlock_rdwait)
if there is an actual conflict.
|
|\ \
| |/ |
|
| |
| |
| |
| |
| |
| |
| | |
The greedy fetch_add(1) approach of read_trylock() may cause
starvation of a waiting write lock request. Let us use a
compare-and-swap for the read lock acquisition in order to
guarantee the progress of writers.
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Do not resend metadata, if metadata does not change between prepare and
execute of prepared statement, or between executes.
Currently, metadata of *every* prepared statement will be checksummed,
and change is detected once checksum changes.
This is not from ideal, performance-wise. The code for
better/faster detection of unchanged metadata, is already in place, but
currently disabled due to PS bugs, such as MDEV-23913.
|
|\ \
| |/ |
|
| |
| |
| |
| |
| |
| | |
Since commit 30ea63b7d2077883713e63cbf4e661ba0345bf68
we actually depend on futex on Linux. Also, we depend on
std::atomic for even longer.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
We always defined PFS_SKIP_BUFFER_MUTEX_RWLOCK, that is,
the latches of the buffer pool blocks were never instrumented
in PERFORMANCE_SCHEMA.
For some reason, the debug_latch (which enforce proper usage of
buffer-fixing in debug builds) was instrumented.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In commit bf3c862faa8efed4a662725ec27586cd69e9228e we introduced
an assertion that may dereference a null pointer.
This regression was caught by running the following:
./mtr --parallel=auto --suite=innodb \
--mysqld=--loose-innodb-adaptive-hash-index
The adaptive hash index is disabled by default since
commit 88cdfc5c7d4764e5267c87eadeb8c3f95faa73d0 (MDEV-20487)
and hence the problem was not caught earlier.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The test seems to deterministically fail on RelWithDebInfo builds
due to a timeout in wait_condition.inc.
According to Matthias Leich (the original author of the test),
the failure rate would reduce if we disabled the purge of
transaction history by setting innodb_force_recovery=2.
For now, let us run this stress test on debug builds only.
|
| |
| |
| |
| |
| |
| |
| |
| | |
fil_space_t::flush_low(): Define and declare without inline.
ut_is_2pow(): Remove UNIV_LIKELY. This is almost exclusively
used in debug assertions. UNIV_LIKELY is not compatible with
static_assert in some compilers.
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
pipeline in community BB
Fix for rebuild from source step
Disable MCS on i386|i686 platforms
This patch puts MCS debian packaging files and part of debian/control
into the engine directory
|
| |
| |
| |
| |
| | |
Centos/RHEL7 have the MAP_HUGE_SHIFT constant
defined in linux/mman.h which needed to get included.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When MDEV-19544 (commit 1a6f470464171bd1144e4dd6f169bb4018f2e81a)
simplified the initialization of the local variable
set_also_gap_locks, an inadvertent change was included.
Essentially, all code branches that are executed when
set_also_gap_locks hold must also ensure that
trx->isolation_level > TRX_ISO_READ_COMMITTED holds.
This was being violated in a few code paths.
It turns out that there is an even simpler fix: Remove the test
of thd_is_select() completely. In that way, the first part of
UPDATE or DELETE should work exactly like SELECT...FOR UPDATE.
thd_is_select(): Remove.
|
| |\ |
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The data member tv_usec of the struct timeval is declared as suseconds_t
on MacOS. Size of suseconds_t is 4 bytes. On the other hand, size of ulong
is 8 bytes on 64-bit MacOS, so attempt to assign a value of wider type
(usec) to a value (tv_usec) of narrower type leads to error.
|
| | |\ |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Starting with commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399)
the function buf_flush_page() will first acquire block->lock and only
after that invoke set_io_fix(). Before that, it was possible to reach
a livelock between buf_page_create() and buf_flush_page().
buf_page_create(): Directly try acquiring the exclusive page latch
without checking whether the page is io-fixed or buffer-fixed.
(As a matter of fact, the have_x_latch() check is not strictly necessary,
because we still support recursive X-latches.)
In case of a latch conflict, wait while allowing buf_page_write_complete()
to acquire buf_pool.mutex and release the block->lock.
An attempt to wait for exclusive block->lock while holding buf_pool.mutex
would lead to a hang in the tests parts.part_supported_sql_func_innodb
and stress.ddl_innodb, due to a deadlock between buf_page_write_complete()
and buf_page_create().
Similarly, in case of an I/O fixed compressed-only
ROW_FORMAT=COMPRESSED page, we will sleep before retrying.
In both cases, we will sleep for 1ms or until a flush batch is completed.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
to 10.5, mysql_upgrade should take of that
Post push fix. Update version to 10.5.8.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This allows MariaDB to compile on old (limits to >2.6.32)
linux kernel versions.
This warns that attempts to use large pages will rely on
implict kernel determination.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This reverts commit 6cf8f05fd9deb900a78898576b85753e09feddaa.
Original patch assumed that MAP_HUGETLB as consistent across
achitectures which isn't the case. Defining it unconditionally
broke large pages on every achitecutre where the value differed
from x86_64.
With the EOL for Centos/RHEL6 announced in 10.5.7, <3.8 linux
kernels are no longer supported.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Account for variety of mips, hppa, solaris and other messages.
Copied from rpl.rpl_drop_db test.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
to 10.5, mysql_upgrade should take of that
Fixing a post push test issue.
|