Implement VATS in XtraDB and InnoDB.

sql/sql_class.cc

@@ -72,6 +72,7 @@
 #include <sys/syscall.h>
 #endif
 
+bool is_slave_replication = false;
 /*
   The following is used to initialise Table_ident with a internal
   table name

sql/sql_class.h

@@ -152,6 +152,8 @@ extern MYSQL_PLUGIN_IMPORT const char **errmesg;
 
 extern bool volatile shutdown_in_progress;
 
+extern bool is_slave_replication;
+
 extern "C" LEX_STRING * thd_query_string (MYSQL_THD thd);
 extern "C" char **thd_query(MYSQL_THD thd);
 

storage/innobase/handler/ha_innodb.cc

@@ -379,6 +379,22 @@ static TYPELIB innodb_default_row_format_typelib = {
 	NULL
 };
 
+/** Possible values of the parameter innodb_lock_schedule_algorithm */
+static const char* innodb_lock_schedule_algorithm_names[] = {
+	"fcfs",
+	"vats",
+	NullS
+};
+
+/** Used to define an enumerate type of the system variable
+innodb_lock_schedule_algorithm. */
+static TYPELIB innodb_lock_schedule_algorithm_typelib = {
+	array_elements(innodb_lock_schedule_algorithm_names) - 1,
+	"innodb_lock_schedule_algorithm_typelib",
+	innodb_lock_schedule_algorithm_names,
+	NULL
+};
+
 /* The following counter is used to convey information to InnoDB
 about server activity: in case of normal DML ops it is not
 sensible to call srv_active_wake_master_thread after each
@@ -22783,6 +22799,18 @@ static MYSQL_SYSVAR_ULONG(doublewrite_batch_size, srv_doublewrite_batch_size,
   NULL, NULL, 120, 1, 127, 0);
 #endif /* defined UNIV_DEBUG || defined UNIV_PERF_DEBUG */
 
+static MYSQL_SYSVAR_ENUM(lock_schedule_algorithm, innodb_lock_schedule_algorithm,
+  PLUGIN_VAR_RQCMDARG,
+  "The algorithm Innodb uses for deciding which locks to grant next when"
+  " a lock is released. Possible values are"
+  " FCFS"
+  " grant the locks in First-Come-First-Served order;"
+  " VATS"
+  " use the Variance-Aware-Transaction-Scheduling algorithm, which"
+  " uses an Eldest-Transaction-First heuristic.",
+  NULL, NULL, INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS,
+  &innodb_lock_schedule_algorithm_typelib);
+
 static MYSQL_SYSVAR_ULONG(buffer_pool_instances, srv_buf_pool_instances,
   PLUGIN_VAR_RQCMDARG | PLUGIN_VAR_READONLY,
   "Number of buffer pool instances, set to higher value on high-end machines to increase scalability",
@@ -23656,6 +23684,7 @@ static struct st_mysql_sys_var* innobase_system_variables[]= {
   MYSQL_SYSVAR(ft_sort_pll_degree),
   MYSQL_SYSVAR(large_prefix),
   MYSQL_SYSVAR(force_load_corrupted),
+  MYSQL_SYSVAR(lock_schedule_algorithm),
   MYSQL_SYSVAR(locks_unsafe_for_binlog),
   MYSQL_SYSVAR(lock_wait_timeout),
   MYSQL_SYSVAR(page_size),

storage/innobase/include/lock0lock.h

@@ -40,6 +40,17 @@ Created 5/7/1996 Heikki Tuuri
 #include "gis0rtree.h"
 #include "lock0prdt.h"
 
+/** Alternatives for innodb_lock_schedule_algorithm, which can be changed by
+    setting innodb_lock_schedule_algorithm. */
+enum innodb_lock_schedule_algorithm_t {
+    /*!< First Come First Served */
+    INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS,
+    /*!< Variance-Aware-Transaction-Scheduling */
+    INNODB_LOCK_SCHEDULE_ALGORITHM_VATS
+};
+
+extern ulong innodb_lock_schedule_algorithm;
+
 // Forward declaration
 class ReadView;
 

storage/xtradb/handler/ha_innodb.cc

@@ -348,6 +348,22 @@ static TYPELIB innodb_empty_free_list_algorithm_typelib = {
 	NULL
 };
 
+/** Possible values of the parameter innodb_lock_schedule_algorithm */
+static const char* innodb_lock_schedule_algorithm_names[] = {
+	"fcfs",
+	"vats",
+	NullS
+};
+
+/** Used to define an enumerate type of the system variable
+innodb_lock_schedule_algorithm. */
+static TYPELIB innodb_lock_schedule_algorithm_typelib = {
+	array_elements(innodb_lock_schedule_algorithm_names) - 1,
+	"innodb_lock_schedule_algorithm_typelib",
+	innodb_lock_schedule_algorithm_names,
+	NULL
+};
+
 /* The following counter is used to convey information to InnoDB
 about server activity: in case of normal DML ops it is not
 sensible to call srv_active_wake_master_thread after each
@@ -20390,6 +20406,18 @@ static MYSQL_SYSVAR_ENUM(empty_free_list_algorithm,
   innodb_srv_empty_free_list_algorithm_validate, NULL, SRV_EMPTY_FREE_LIST_BACKOFF,
   &innodb_empty_free_list_algorithm_typelib);
 
+static MYSQL_SYSVAR_ENUM(lock_schedule_algorithm, innodb_lock_schedule_algorithm,
+  PLUGIN_VAR_RQCMDARG,
+  "The algorithm Innodb uses for deciding which locks to grant next when"
+  " a lock is released. Possible values are"
+  " FCFS"
+  " grant the locks in First-Come-First-Served order;"
+  " VATS"
+  " use the Variance-Aware-Transaction-Scheduling algorithm, which"
+  " uses an Eldest-Transaction-First heuristic.",
+  NULL, NULL, INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS,
+  &innodb_lock_schedule_algorithm_typelib);
+
 static MYSQL_SYSVAR_LONG(buffer_pool_instances, innobase_buffer_pool_instances,
   PLUGIN_VAR_RQCMDARG | PLUGIN_VAR_READONLY,
   "Number of buffer pool instances, set to higher value on high-end machines to increase scalability",
@@ -21283,6 +21311,7 @@ static struct st_mysql_sys_var* innobase_system_variables[]= {
   MYSQL_SYSVAR(ft_sort_pll_degree),
   MYSQL_SYSVAR(large_prefix),
   MYSQL_SYSVAR(force_load_corrupted),
+  MYSQL_SYSVAR(lock_schedule_algorithm),
   MYSQL_SYSVAR(locks_unsafe_for_binlog),
   MYSQL_SYSVAR(lock_wait_timeout),
 #ifdef UNIV_LOG_ARCHIVE

storage/xtradb/include/lock0lock.h

@@ -47,6 +47,17 @@ extern ibool	lock_print_waits;
 
 extern ulint	srv_n_lock_deadlock_count;
 
+/** Alternatives for innodb_lock_schedule_algorithm, which can be changed by
+    setting innodb_lock_schedule_algorithm. */
+enum innodb_lock_schedule_algorithm_t {
+    /*!< First Come First Served */
+    INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS,
+    /*!< Variance-Aware-Transaction-Scheduling */
+    INNODB_LOCK_SCHEDULE_ALGORITHM_VATS
+};
+
+extern ulong innodb_lock_schedule_algorithm;
+
 /*********************************************************************//**
 Gets the size of a lock struct.
 @return	size in bytes */

storage/xtradb/lock/lock0lock.cc

@@ -385,6 +385,11 @@ extern "C" int thd_need_wait_reports(const MYSQL_THD thd);
 extern "C"
 int thd_need_ordering_with(const MYSQL_THD thd, const MYSQL_THD other_thd);
 
+/** Lock scheduling algorithm */
+ulong innodb_lock_schedule_algorithm = INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS;
+
+extern "C" int thd_deadlock_victim_preference(const MYSQL_THD thd1, const MYSQL_THD thd2);
+
 /** Stack to use during DFS search. Currently only a single stack is required
 because there is no parallel deadlock check. This stack is protected by
 the lock_sys_t::mutex. */
@@ -2198,6 +2203,85 @@ lock_rec_create(
 	return(lock);
 }
 
+/*********************************************************************//**
+Check if lock1 has higher priority than lock2.
+NULL has lowest priority.
+If either is a high priority transaction, the lock has higher priority.
+If neither of them is wait lock, the first one has higher priority.
+If only one of them is a wait lock, it has lower priority.
+Otherwise, the one with an older transaction has higher priority.
+@returns true if lock1 has higher priority, false otherwise. */
+bool
+has_higher_priority(
+	lock_t *lock1,
+	lock_t *lock2)
+{
+	if (lock1 == NULL) {
+		return false;
+	} else if (lock2 == NULL) {
+		return true;
+	}
+	// Ask the upper server layer if any of the two trx should be prefered.
+	int preference = thd_deadlock_victim_preference(lock1->trx->mysql_thd, lock2->trx->mysql_thd);
+	if (preference == -1) {
+		// lock1 is preferred as a victim, so lock2 has higher priority
+		return false;
+	} else if (preference == 1) {
+		// lock2 is preferred as a victim, so lock1 has higher priority
+		return true;
+	}
+	// No preference. Compre them by wait mode and trx age.
+	if (!lock_get_wait(lock1)) {
+		return true;
+	} else if (!lock_get_wait(lock2)) {
+		return false;
+	}
+	return lock1->trx->start_time < lock2->trx->start_time;
+}
+
+/*********************************************************************//**
+Insert a lock to the hash list according to the mode (whether it is a wait
+lock) and the age of the transaction the it is associated with.
+If the lock is not a wait lock, insert it to the head of the hash list.
+Otherwise, insert it to the middle of the wait locks according to the age of
+the transaciton. */
+static
+void
+lock_rec_insert_by_trx_age(
+	lock_t *in_lock, /*!< in: lock to be insert */
+	bool wait)	 /*!< in: whether it's a wait lock */
+{
+    ulint               space;
+    ulint               page_no;
+	ulint				rec_fold;
+    hash_table_t*       hash;
+	hash_cell_t*        cell;
+	lock_t*				node;
+	lock_t*				next;
+
+    space = in_lock->un_member.rec_lock.space;
+    page_no = in_lock->un_member.rec_lock.page_no;
+	rec_fold = lock_rec_fold(space, page_no);
+    hash = lock_hash_get(in_lock->type_mode);
+	cell = hash_get_nth_cell(hash,
+				 hash_calc_hash(rec_fold, hash));
+
+	node = (lock_t *) cell->node;
+	// If in_lock is not a wait lock, we insert it to the head of the list.
+	if (node == NULL || !wait || has_higher_priority(in_lock, node)) {
+		cell->node = in_lock;
+		in_lock->hash = node;
+		return;
+	}
+	while (node != NULL && has_higher_priority((lock_t *) node->hash,
+						   in_lock)) {
+		node = (lock_t *) node->hash;
+	}
+	next = (lock_t *) node->hash;
+	node->hash = in_lock;
+    in_lock->hash = next;
+}
+
 /*********************************************************************//**
 Enqueues a waiting request for a lock which cannot be granted immediately.
 Checks for deadlocks.
@@ -2300,7 +2384,23 @@ lock_rec_enqueue_waiting(
 
 		return(DB_DEADLOCK);
 
-	} else if (trx->lock.wait_lock == NULL) {
+	}
+
+    // Move it only when it does not cause a deadlock.
+    if (innodb_lock_schedule_algorithm
+        == INNODB_LOCK_SCHEDULE_ALGORITHM_VATS
+        && !trx_is_high_priority(lock->trx)) {
+        
+        HASH_DELETE(lock_t, hash, lock_hash_get(lock->type_mode),
+                    m_rec_id.fold(), lock);
+        lock_rec_insert_by_trx_age(lock, m_mode & LOCK_WAIT);
+        if (lock_get_wait(lock) && !lock_rec_has_to_wait_in_queue(lock)) {
+            lock_reset_lock_and_trx_wait(lock);
+            return DB_SUCCESS_LOCKED_REC;
+        }
+    }
+
+	if (trx->lock.wait_lock == NULL) {
 
 		/* If there was a deadlock but we chose another
 		transaction as a victim, it is possible that we
@@ -2858,6 +2958,32 @@ lock_rec_cancel(
 	trx_mutex_exit(lock->trx);
 }
 
+/*************************************************************//**
+Move the lock to the head of the hash list. */
+static
+void
+lock_rec_move_to_front(
+	lock_t *lock_to_move,	/*!< in: lock to be moved */
+	ulint rec_fold)			 /*!< in: rec fold of the lock */
+{
+    hash_table_t*	lock_hash;
+    hash_cell_t*    cell;
+    lock_t*         next;
+    
+	if (lock_to_move != NULL)
+	{
+        lock_hash = lock_hash_get(lock_to_move->type_mode);
+		// Move the target lock to the head of the list
+		cell = hash_get_nth_cell(lock_hash,
+                hash_calc_hash(rec_fold, lock_hash));
+		if (lock_to_move != cell->node) {
+			next = (lock_t *) cell->node;
+			cell->node = lock_to_move;
+			lock_to_move->hash = next;
+		}
+	}
+}
+
 /*************************************************************//**
 Removes a record lock request, waiting or granted, from the queue and
 grants locks to other transactions in the queue if they now are entitled
@@ -2875,6 +3001,8 @@ lock_rec_dequeue_from_page(
 {
 	ulint		space;
 	ulint		page_no;
+	ulint		rec_fold;
+	lock_t*		previous = NULL;
 	lock_t*		lock;
 	trx_lock_t*	trx_lock;
 
@@ -2886,6 +3014,7 @@ lock_rec_dequeue_from_page(
 
 	space = in_lock->un_member.rec_lock.space;
 	page_no = in_lock->un_member.rec_lock.page_no;
+	rec_fold = lock_rec_fold(space, page_no);
 
 	in_lock->index->table->n_rec_locks--;
 
@@ -2898,20 +3027,58 @@ lock_rec_dequeue_from_page(
 	MONITOR_INC(MONITOR_RECLOCK_REMOVED);
 	MONITOR_DEC(MONITOR_NUM_RECLOCK);
 
-	/* Check if waiting locks in the queue can now be granted: grant
-	locks if there are no conflicting locks ahead. Stop at the first
-	X lock that is waiting or has been granted. */
+	if (innodb_lock_schedule_algorithm
+	    == INNODB_LOCK_SCHEDULE_ALGORITHM_FCFS) {
 
-	for (lock = lock_rec_get_first_on_page_addr(space, page_no);
-	     lock != NULL;
-	     lock = lock_rec_get_next_on_page(lock)) {
+		/* Check if waiting locks in the queue can now be granted:
+		grant locks if there are no conflicting locks ahead. Stop at
+		the first X lock that is waiting or has been granted. */
 
-		if (lock_get_wait(lock)
-		    && !lock_rec_has_to_wait_in_queue(lock)) {
+		for (lock = lock_rec_get_first_on_page_addr(lock_hash, space,
+							    page_no);
+		     lock != NULL;
+		     lock = lock_rec_get_next_on_page(lock)) {
 
-			/* Grant the lock */
-			ut_ad(lock->trx != in_lock->trx);
-			lock_grant(lock);
+			if (lock_get_wait(lock)
+			    && !lock_rec_has_to_wait_in_queue(lock)) {
+
+				/* Grant the lock */
+				ut_ad(lock->trx != in_lock->trx);
+
+				lock_grant(lock);
+			}
+		}
+	} else {
+		/* Grant locks if there are no conflicting locks ahead.
+		Move granted locks to the head of the list. */
+		for (lock = lock_rec_get_first_on_page_addr(lock_hash, space, page_no);
+		     lock != NULL;) {
+
+			/* If the lock is a wait lock on this page, and it does not need to wait. */
+			if ((lock->un_member.rec_lock.space == space)
+			    && (lock->un_member.rec_lock.page_no == page_no)
+			    && lock_get_wait(lock)
+			    && !lock_rec_has_to_wait_in_queue(lock)) {
+
+				/* Grant the lock */
+				ut_ad(lock->trx != in_lock->trx);
+
+				lock_grant(lock);
+
+				if (previous != NULL) {
+					/* Move the lock to the head of the list. */
+					HASH_GET_NEXT(hash, previous) = HASH_GET_NEXT(hash, lock);
+					lock_rec_move_to_front(lock, rec_fold);
+				} else {
+					/* Already at the head of the list. */
+					previous = lock;
+				}
+				/* Move on to the next lock. */
+				lock = static_cast<lock_t *>(HASH_GET_NEXT(hash, previous));
+			} else {
+				previous = lock;
+				lock = static_cast<lock_t *>(HASH_GET_NEXT(hash, lock));
+			}
 		}
 	}
 }
@@ -4118,7 +4285,8 @@ lock_get_first_lock(
 	}
 
 	ut_a(lock != NULL);
-	ut_a(lock != ctx->wait_lock);
+	ut_a(lock != ctx->wait_lock ||
+         innodb_lock_schedule_algorithm == INNODB_LOCK_SCHEDULE_ALGORITHM_VATS);
 	ut_ad(lock_get_type_low(lock) == lock_get_type_low(ctx->wait_lock));
 
 	return(lock);