MDEV-25114 Crash: WSREP: invalid state ROLLED_BACK (FATAL)bb-10.4-MDEV-25114

This patch is the plan D variant for fixing potetial mutex locking
order exercised by BF aborting and KILL command execution.

In this approach, KILL command is replicated as TOI operation.
This guarantees total isolation for the KILL command execution
in the first node: there is no concurrent replication applying
and no concurrent DDL executing. Therefore there is no risk of
BF aborting to happen in parallel with KILL command execution
either. Potential mutex deadlocks between the different mutex
access paths with KILL command execution and BF aborting cannot
therefore happen.

TOI replication is used, in this approach,  purely as means
to provide isolated KILL command execution in the first node.
KILL command should not (and must not) be applied in secondary
nodes. In this patch, we make this sure by skipping KILL
execution in secondary nodes, in applying phase, where we
bail out if applier thread is trying to execute KILL command.
This is effective, but skipping the applying of KILL command
could happen much earlier as well.

This patch also fixes mutex locking order and unprotected
THD member accesses on bf aborting case. We try to hold
THD::LOCK_thd_data during bf aborting. Only case where it
is not possible is at wsrep_abort_transaction before
call wsrep_innobase_kill_one_trx where we take InnoDB
mutexes first and then THD::LOCK_thd_data.

This will also fix possible race condition during
close_connection and while wsrep is disconnecting
connections.

Added wsrep_bf_kill_debug test case

Reviewed-by: Jan Lindström <jan.lindstrom@mariadb.com>

1 files changed, 18 insertions, 8 deletions

diff --git a/sql/sql_class.cc b/sql/sql_class.cc
index 28bf77c94e8..c2dcc07997e 100644
--- a/sql/sql_class.cc
+++ b/sql/sql_class.cc
@@ -1977,13 +1977,13 @@ void THD::abort_current_cond_wait(bool force)
           the Vio might be disassociated concurrently.
 */
 
-void THD::disconnect()
+void THD::disconnect_mutexed()
 {
   Vio *vio= NULL;
 
-  set_killed(KILL_CONNECTION);
-
-  mysql_mutex_lock(&LOCK_thd_data);
+  mysql_mutex_assert_owner(&LOCK_thd_data);
+  mysql_mutex_assert_owner(&LOCK_thd_kill);
+  set_killed_no_mutex(KILL_CONNECTION);
 
 #ifdef SIGNAL_WITH_VIO_CLOSE
   /*
@@ -1999,8 +1999,6 @@ void THD::disconnect()
   if (net.vio != vio)
     vio_close(net.vio);
   net.thd= 0;                                   // Don't collect statistics
-
-  mysql_mutex_unlock(&LOCK_thd_data);
 }
 
 
@@ -2027,6 +2025,9 @@ bool THD::notify_shared_lock(MDL_context_owner *ctx_in_use,
 
   if (needs_thr_lock_abort)
   {
+    /* Protect thread from concurrent disconnect and delete */
+    mysql_mutex_lock(&in_use->LOCK_thd_kill);
+    /* Protect thread from concurrent usage */
     mysql_mutex_lock(&in_use->LOCK_thd_data);
     /* If not already dying */
     if (in_use->killed != KILL_CONNECTION_HARD)
@@ -2043,11 +2044,20 @@ bool THD::notify_shared_lock(MDL_context_owner *ctx_in_use,
           thread can see those instances (e.g. see partitioning code).
         */
         if (!thd_table->needs_reopen())
-        {
           signalled|= mysql_lock_abort_for_thread(this, thd_table);
-        }
       }
+#ifdef WITH_WSREP
+      if (WSREP(this) && wsrep_thd_is_BF(this, false))
+      {
+        WSREP_DEBUG("notify_shared_lock: BF thread %llu query %s"
+                    " victim %llu query %s",
+                    this->real_id, wsrep_thd_query(this),
+                    in_use->real_id, wsrep_thd_query(in_use));
+        wsrep_abort_thd(this, in_use, false);
+      }
+#endif /* WITH_WSREP */
     }
+    mysql_mutex_unlock(&in_use->LOCK_thd_kill);
     mysql_mutex_unlock(&in_use->LOCK_thd_data);
   }
   DBUG_RETURN(signalled);