diff options
| author | Marko Mäkelä <marko.makela@mariadb.com> | 2017-02-16 19:40:03 +0200 |
|---|---|---|
| committer | Marko Mäkelä <marko.makela@mariadb.com> | 2017-02-20 12:20:52 +0200 |
| commit | 13493078e9aea37f6714b3921921aa10864c8b30 (patch) | |
| tree | 904582d100122e966de83891ce372908d128294d /storage/innobase/include/lock0lock.h | |
| parent | 72994d6442ec6156ccfefb260771537a8b6d80b9 (diff) | |
| download | mariadb-git-13493078e9aea37f6714b3921921aa10864c8b30.tar.gz | |
MDEV-11802 innodb.innodb_bug14676111 fails
The function trx_purge_stop() was calling os_event_reset(purge_sys->event)
before calling rw_lock_x_lock(&purge_sys->latch). The os_event_set()
call in srv_purge_coordinator_suspend() is protected by that X-latch.
It would seem a good idea to consistently protect both os_event_set()
and os_event_reset() calls with a common mutex or rw-lock in those
cases where os_event_set() and os_event_reset() are used
like condition variables, tied to changes of shared state.
For each os_event_t, we try to document the mutex or rw-lock that is
being used. For some events, frequent calls to os_event_set() seem to
try to avoid hangs. Some events are never waited for infinitely, only
timed waits, and os_event_set() is used for early termination of these
waits.
os_aio_simulated_put_read_threads_to_sleep(): Define as a null macro
on other systems than Windows. TODO: remove this altogether and disable
innodb_use_native_aio on Windows.
os_aio_segment_wait_events[]: Initialize only if innodb_use_native_aio=0.
Diffstat (limited to 'storage/innobase/include/lock0lock.h')
| -rw-r--r-- | storage/innobase/include/lock0lock.h | 16 |
1 files changed, 11 insertions, 5 deletions
diff --git a/storage/innobase/include/lock0lock.h b/storage/innobase/include/lock0lock.h index 4bfc60690ce..7fbc19086d5 100644 --- a/storage/innobase/include/lock0lock.h +++ b/storage/innobase/include/lock0lock.h @@ -918,7 +918,12 @@ struct lock_sys_t{ srv_slot_t* waiting_threads; /*!< Array of user threads suspended while waiting for locks within InnoDB, protected - by the lock_sys->wait_mutex */ + by the lock_sys->wait_mutex; + os_event_set() and + os_event_reset() on + waiting_threads[]->event + are protected by + trx_t::mutex */ srv_slot_t* last_slot; /*!< highest slot ever used in the waiting_threads array, protected by @@ -931,10 +936,11 @@ struct lock_sys_t{ ulint n_lock_max_wait_time; /*!< Max wait time */ - os_event_t timeout_event; /*!< Set to the event that is - created in the lock wait monitor - thread. A value of 0 means the - thread is not active */ + os_event_t timeout_event; /*!< An event waited for by + lock_wait_timeout_thread. + Not protected by a mutex, + but the waits are timed. + Signaled on shutdown only. */ bool timeout_thread_active; /*!< True if the timeout thread is running */ |