mariadb/sql/wsrep_thd.cc
Marko Mäkelä d0116e10a5 Revert MDEV-18464 and MDEV-12009
This reverts commit 21b2fada7a
and commit 81d71ee6b2.

The MDEV-18464 change introduces a few data race issues. Contrary to
the documentation, the field trx_t::victim is not always being protected
by lock_sys_t::mutex and trx_t::mutex. Most importantly, it seems
that KILL QUERY could wrongly avoid acquiring both mutexes when
invoking lock_trx_handle_wait_low(), in case another thread had
already set trx->victim=true.

We also revert MDEV-12009, because it should depend on the MDEV-18464
fix being present.
2019-03-28 12:39:50 +02:00

417 lines
14 KiB
C++

/* Copyright (C) 2013 Codership Oy <info@codership.com>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. */
#include "mariadb.h"
#include "wsrep_thd.h"
#include "wsrep_trans_observer.h"
#include "wsrep_high_priority_service.h"
#include "wsrep_storage_service.h"
#include "transaction.h"
#include "rpl_rli.h"
#include "log_event.h"
#include "sql_parse.h"
#include "sql_base.h" // close_thread_tables()
#include "mysqld.h" // start_wsrep_THD();
#include "wsrep_applier.h" // start_wsrep_THD();
#include "mysql/service_wsrep.h"
#include "debug_sync.h"
#include "slave.h"
#include "rpl_rli.h"
#include "rpl_mi.h"
static Wsrep_thd_queue* wsrep_rollback_queue= 0;
static Wsrep_thd_queue* wsrep_post_rollback_queue= 0;
static Atomic_counter<uint64_t> wsrep_bf_aborts_counter;
int wsrep_show_bf_aborts (THD *thd, SHOW_VAR *var, char *buff,
enum enum_var_type scope)
{
wsrep_local_bf_aborts= wsrep_bf_aborts_counter;
var->type= SHOW_LONGLONG;
var->value= (char*)&wsrep_local_bf_aborts;
return 0;
}
static void wsrep_replication_process(THD *thd,
void* arg __attribute__((unused)))
{
DBUG_ENTER("wsrep_replication_process");
Wsrep_applier_service applier_service(thd);
/* thd->system_thread_info.rpl_sql_info isn't initialized. */
thd->system_thread_info.rpl_sql_info=
new rpl_sql_thread_info(thd->wsrep_rgi->rli->mi->rpl_filter);
WSREP_INFO("Starting applier thread %llu", thd->thread_id);
enum wsrep::provider::status
ret= Wsrep_server_state::get_provider().run_applier(&applier_service);
WSREP_INFO("Applier thread exiting ret: %d thd: %llu", ret, thd->thread_id);
mysql_mutex_lock(&LOCK_wsrep_slave_threads);
wsrep_close_applier(thd);
mysql_cond_broadcast(&COND_wsrep_slave_threads);
mysql_mutex_unlock(&LOCK_wsrep_slave_threads);
delete thd->system_thread_info.rpl_sql_info;
delete thd->wsrep_rgi->rli->mi;
delete thd->wsrep_rgi->rli;
thd->wsrep_rgi->cleanup_after_session();
delete thd->wsrep_rgi;
thd->wsrep_rgi= NULL;
if(thd->has_thd_temporary_tables())
{
WSREP_WARN("Applier %lld has temporary tables at exit.",
thd->thread_id);
}
DBUG_VOID_RETURN;
}
static bool create_wsrep_THD(Wsrep_thd_args* args)
{
ulong old_wsrep_running_threads= wsrep_running_threads;
pthread_t unused;
bool res= pthread_create(&unused, &connection_attrib, start_wsrep_THD,
args);
/*
if starting a thread on server startup, wait until the this thread's THD
is fully initialized (otherwise a THD initialization code might
try to access a partially initialized server data structure - MDEV-8208).
*/
mysql_mutex_lock(&LOCK_wsrep_slave_threads);
if (!mysqld_server_initialized)
while (old_wsrep_running_threads == wsrep_running_threads)
mysql_cond_wait(&COND_wsrep_slave_threads, &LOCK_wsrep_slave_threads);
mysql_mutex_unlock(&LOCK_wsrep_slave_threads);
return res;
}
void wsrep_create_appliers(long threads)
{
/* Dont' start slave threads if wsrep-provider or wsrep-cluster-address
is not set.
*/
if (!WSREP_PROVIDER_EXISTS)
{
return;
}
if (!wsrep_cluster_address || wsrep_cluster_address[0]== 0)
{
WSREP_DEBUG("wsrep_create_appliers exit due to empty address");
return;
}
long wsrep_threads=0;
while (wsrep_threads++ < threads)
{
Wsrep_thd_args* args(new Wsrep_thd_args(wsrep_replication_process, 0));
if (create_wsrep_THD(args))
{
WSREP_WARN("Can't create thread to manage wsrep replication");
}
}
}
static void wsrep_rollback_process(THD *rollbacker,
void *arg __attribute__((unused)))
{
DBUG_ENTER("wsrep_rollback_process");
THD* thd= NULL;
DBUG_ASSERT(!wsrep_rollback_queue);
wsrep_rollback_queue= new Wsrep_thd_queue(rollbacker);
WSREP_INFO("Starting rollbacker thread %llu", rollbacker->thread_id);
thd_proc_info(rollbacker, "wsrep aborter idle");
while ((thd= wsrep_rollback_queue->pop_front()) != NULL)
{
mysql_mutex_lock(&thd->LOCK_thd_data);
wsrep::client_state& cs(thd->wsrep_cs());
const wsrep::transaction& tx(cs.transaction());
if (tx.state() == wsrep::transaction::s_aborted)
{
WSREP_DEBUG("rollbacker thd already aborted: %llu state: %d",
(long long)thd->real_id,
tx.state());
mysql_mutex_unlock(&thd->LOCK_thd_data);
continue;
}
mysql_mutex_unlock(&thd->LOCK_thd_data);
thd_proc_info(rollbacker, "wsrep aborter active");
wsrep::transaction_id transaction_id(thd->wsrep_trx().id());
if (thd->wsrep_trx().is_streaming() &&
thd->wsrep_trx().bf_aborted_in_total_order())
{
thd->store_globals();
thd->wsrep_cs().store_globals();
if (thd->wsrep_cs().mode() == wsrep::client_state::m_high_priority)
{
DBUG_ASSERT(thd->wsrep_applier_service);
thd->wsrep_applier_service->rollback(wsrep::ws_handle(),
wsrep::ws_meta());
thd->wsrep_applier_service->after_apply();
/* Will free THD */
Wsrep_server_state::instance().server_service().
release_high_priority_service(thd->wsrep_applier_service);
}
else
{
mysql_mutex_lock(&thd->LOCK_thd_data);
/* prepare THD for rollback processing */
thd->reset_for_next_command(true);
thd->lex->sql_command= SQLCOM_ROLLBACK;
mysql_mutex_unlock(&thd->LOCK_thd_data);
/* Perform a client rollback, restore globals and signal
the victim only when all the resources have been
released */
thd->wsrep_cs().client_service().bf_rollback();
thd->reset_globals();
thd->wsrep_cs().sync_rollback_complete();
}
}
else if (wsrep_thd_is_applying(thd))
{
WSREP_DEBUG("rollbacker aborting SR thd: (%lld %llu)",
thd->thread_id, (long long)thd->real_id);
DBUG_ASSERT(thd->wsrep_cs().mode() == Wsrep_client_state::m_high_priority);
/* Must be streaming and must have been removed from the
server state streaming appliers map. */
DBUG_ASSERT(thd->wsrep_trx().is_streaming());
DBUG_ASSERT(!Wsrep_server_state::instance().find_streaming_applier(
thd->wsrep_trx().server_id(),
thd->wsrep_trx().id()));
DBUG_ASSERT(thd->wsrep_applier_service);
/* Fragment removal should happen before rollback to make
the transaction non-observable in SR table after the rollback
completes. For correctness the order does not matter here,
but currently it is mandated by checks in some MTR tests. */
Wsrep_storage_service* storage_service=
static_cast<Wsrep_storage_service*>(
Wsrep_server_state::instance().server_service().storage_service(
*thd->wsrep_applier_service));
storage_service->store_globals();
storage_service->adopt_transaction(thd->wsrep_trx());
storage_service->remove_fragments();
storage_service->commit(wsrep::ws_handle(transaction_id, 0),
wsrep::ws_meta());
Wsrep_server_state::instance().server_service().release_storage_service(storage_service);
thd->store_globals();
thd->wsrep_cs().store_globals();
thd->wsrep_applier_service->rollback(wsrep::ws_handle(),
wsrep::ws_meta());
thd->wsrep_applier_service->after_apply();
/* Will free THD */
Wsrep_server_state::instance().server_service()
.release_high_priority_service(thd->wsrep_applier_service);
}
else
{
if (thd->wsrep_trx().is_streaming())
{
Wsrep_storage_service* storage_service=
static_cast<Wsrep_storage_service*>(
Wsrep_server_state::instance().server_service().
storage_service(thd->wsrep_cs().client_service()));
storage_service->store_globals();
storage_service->adopt_transaction(thd->wsrep_trx());
storage_service->remove_fragments();
storage_service->commit(wsrep::ws_handle(transaction_id, 0),
wsrep::ws_meta());
Wsrep_server_state::instance().server_service().
release_storage_service(storage_service);
}
thd->store_globals();
thd->wsrep_cs().store_globals();
mysql_mutex_lock(&thd->LOCK_thd_data);
/* prepare THD for rollback processing */
thd->reset_for_next_command();
thd->lex->sql_command= SQLCOM_ROLLBACK;
mysql_mutex_unlock(&thd->LOCK_thd_data);
/* Perform a client rollback, restore globals and signal
the victim only when all the resources have been
released */
thd->wsrep_cs().client_service().bf_rollback();
thd->reset_globals();
thd->wsrep_cs().sync_rollback_complete();
WSREP_DEBUG("rollbacker aborted thd: (%llu %llu)",
thd->thread_id, (long long)thd->real_id);
}
thd_proc_info(rollbacker, "wsrep aborter idle");
}
delete wsrep_rollback_queue;
wsrep_rollback_queue= NULL;
WSREP_INFO("rollbacker thread exiting %llu", rollbacker->thread_id);
DBUG_ASSERT(rollbacker->killed != NOT_KILLED);
DBUG_PRINT("wsrep",("wsrep rollbacker thread exiting"));
DBUG_VOID_RETURN;
}
static void wsrep_post_rollback_process(THD *post_rollbacker,
void *arg __attribute__((unused)))
{
DBUG_ENTER("wsrep_post_rollback_process");
THD* thd= NULL;
WSREP_INFO("Starting post rollbacker thread %llu", post_rollbacker->thread_id);
DBUG_ASSERT(!wsrep_post_rollback_queue);
wsrep_post_rollback_queue= new Wsrep_thd_queue(post_rollbacker);
while ((thd= wsrep_post_rollback_queue->pop_front()) != NULL)
{
thd->store_globals();
wsrep::client_state& cs(thd->wsrep_cs());
mysql_mutex_lock(&thd->LOCK_thd_data);
DBUG_ASSERT(thd->wsrep_trx().state() == wsrep::transaction::s_aborting);
WSREP_DEBUG("post rollbacker calling post rollback for thd %llu, conf %s",
thd->thread_id, wsrep_thd_transaction_state_str(thd));
cs.after_rollback();
DBUG_ASSERT(thd->wsrep_trx().state() == wsrep::transaction::s_aborted);
mysql_mutex_unlock(&thd->LOCK_thd_data);
}
delete wsrep_post_rollback_queue;
wsrep_post_rollback_queue= NULL;
DBUG_ASSERT(post_rollbacker->killed != NOT_KILLED);
DBUG_PRINT("wsrep",("wsrep post rollbacker thread exiting"));
WSREP_INFO("post rollbacker thread exiting %llu", post_rollbacker->thread_id);
DBUG_VOID_RETURN;
}
void wsrep_create_rollbacker()
{
if (wsrep_cluster_address && wsrep_cluster_address[0] != 0)
{
Wsrep_thd_args* args= new Wsrep_thd_args(wsrep_rollback_process, 0);
/* create rollbacker */
if (create_wsrep_THD(args))
WSREP_WARN("Can't create thread to manage wsrep rollback");
/* create post_rollbacker */
args= new Wsrep_thd_args(wsrep_post_rollback_process, 0);
if (create_wsrep_THD(args))
WSREP_WARN("Can't create thread to manage wsrep post rollback");
}
}
/*
Start async rollback process
Asserts thd->LOCK_thd_data ownership
*/
void wsrep_fire_rollbacker(THD *thd)
{
DBUG_ASSERT(thd->wsrep_trx().state() == wsrep::transaction::s_aborting);
DBUG_PRINT("wsrep",("enqueuing trx abort for %llu", thd->thread_id));
WSREP_DEBUG("enqueuing trx abort for (%llu)", thd->thread_id);
if (wsrep_rollback_queue->push_back(thd))
{
WSREP_WARN("duplicate thd %llu for rollbacker",
thd->thread_id);
}
}
int wsrep_abort_thd(void *bf_thd_ptr, void *victim_thd_ptr, my_bool signal)
{
DBUG_ENTER("wsrep_abort_thd");
THD *victim_thd= (THD *) victim_thd_ptr;
THD *bf_thd= (THD *) bf_thd_ptr;
mysql_mutex_lock(&victim_thd->LOCK_thd_data);
if ( (WSREP(bf_thd) ||
( (WSREP_ON || bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU) &&
wsrep_thd_is_toi(bf_thd)) ) &&
victim_thd &&
!wsrep_thd_is_aborting(victim_thd))
{
WSREP_DEBUG("wsrep_abort_thd, by: %llu, victim: %llu", (bf_thd) ?
(long long)bf_thd->real_id : 0, (long long)victim_thd->real_id);
mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
ha_abort_transaction(bf_thd, victim_thd, signal);
mysql_mutex_lock(&victim_thd->LOCK_thd_data);
}
else
{
WSREP_DEBUG("wsrep_abort_thd not effective: %p %p", bf_thd, victim_thd);
}
mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
DBUG_RETURN(1);
}
bool wsrep_bf_abort(const THD* bf_thd, THD* victim_thd)
{
WSREP_LOG_THD((THD*)bf_thd, "BF aborter before");
WSREP_LOG_THD(victim_thd, "victim before");
wsrep::seqno bf_seqno(bf_thd->wsrep_trx().ws_meta().seqno());
if (WSREP(victim_thd) && !victim_thd->wsrep_trx().active())
{
WSREP_DEBUG("wsrep_bf_abort, BF abort for non active transaction");
wsrep_start_transaction(victim_thd, victim_thd->wsrep_next_trx_id());
}
bool ret;
if (wsrep_thd_is_toi(bf_thd))
{
ret= victim_thd->wsrep_cs().total_order_bf_abort(bf_seqno);
}
else
{
ret= victim_thd->wsrep_cs().bf_abort(bf_seqno);
}
if (ret)
{
wsrep_bf_aborts_counter++;
}
return ret;
}
/*
Get auto increment variables for THD. Use global settings for
applier threads.
*/
void wsrep_thd_auto_increment_variables(THD* thd,
unsigned long long* offset,
unsigned long long* increment)
{
if (wsrep_thd_is_applying(thd) &&
thd->wsrep_trx().state() != wsrep::transaction::s_replaying)
{
*offset= global_system_variables.auto_increment_offset;
*increment= global_system_variables.auto_increment_increment;
return;
}
*offset= thd->variables.auto_increment_offset;
*increment= thd->variables.auto_increment_increment;
}