MDEV-29433 innodb.lock_delete_updated is unstable

Use suspend thread syncpoint instead of include/wait_condition.inc to
make sure DELETE created waiting lock before the next UPDATE begins
locking.

This is backport of commit 0fa4dd07
from 10.6.

mysql-test/suite/innodb/r/lock_delete_updated.result

 CREATE TABLE t(a INT PRIMARY KEY) ENGINE=InnoDB;
 INSERT INTO t VALUES (3);
 BEGIN;
-connection default;
 UPDATE t SET a = 2;
 connect con1,localhost,root;
+SET DEBUG_SYNC="lock_wait_suspend_thread_enter SIGNAL del_locked";
 DELETE FROM t;
 connection default;
+SET DEBUG_SYNC="now WAIT_FOR del_locked";
 UPDATE t SET a = 1;
 COMMIT;
 connection con1;
@@ -17,4 +18,5 @@ connection default;
 SELECT count(*) FROM t;
 count(*)
 1
+SET DEBUG_SYNC="reset";
 DROP TABLE t;

mysql-test/suite/innodb/t/lock_delete_updated.test

 --source include/have_innodb.inc
 --source include/count_sessions.inc
+--source include/have_debug.inc
+--source include/have_debug_sync.inc
 
 CREATE TABLE t(a INT PRIMARY KEY) ENGINE=InnoDB;
 INSERT INTO t VALUES (3);
 
 BEGIN;
-
-connection default;
 UPDATE t SET a = 2;
 
 connect con1,localhost,root;
+SET DEBUG_SYNC="lock_wait_suspend_thread_enter SIGNAL del_locked";
 send DELETE FROM t;
 
 connection default;
-let $wait_condition=
-    select count(*) = 1 from information_schema.processlist
-    where state = "Updating" and info = "DELETE FROM t";
---source include/wait_condition.inc
-
+SET DEBUG_SYNC="now WAIT_FOR del_locked";
 UPDATE t SET a = 1;
 COMMIT;
 
@@ -30,5 +27,6 @@ connection default;
 --echo # The above DELETE must delete all the rows in the table, so the
 --echo # following SELECT must show 0 rows.
 SELECT count(*) FROM t;
+SET DEBUG_SYNC="reset";
 DROP TABLE t;
 --source include/wait_until_count_sessions.inc