The following test case causes a crash of the server in MariDB 5.2/5.3/5.5 on the second execution of the prepared statement:

CREATE TABLE t1 (
pk INTEGER AUTO_INCREMENT,
col_int_nokey INTEGER,
col_int_key INTEGER,

col_varchar_key VARCHAR(1),
col_varchar_nokey VARCHAR(1),

PRIMARY KEY (pk),
KEY (col_int_key),
KEY (col_varchar_key, col_int_key)
);

INSERT INTO t1 (
col_int_key, col_int_nokey,
col_varchar_key, col_varchar_nokey
) VALUES
(4, 2, 'v', 'v'),
(62, 150, 'v', 'v');

CREATE TABLE t2 (
pk INTEGER AUTO_INCREMENT,
col_int_nokey INTEGER,
col_int_key INTEGER,

col_varchar_key VARCHAR(1),
col_varchar_nokey VARCHAR(1),

PRIMARY KEY (pk),
KEY (col_int_key),
KEY (col_varchar_key, col_int_key)
);

INSERT INTO t2 (
col_int_key, col_int_nokey,
col_varchar_key, col_varchar_nokey
) VALUES
(8, NULL, 'x', 'x'),
(7, 8, 'd', 'd');

PREPARE stmt FROM '
SELECT
( SELECT MAX( SQ1_alias2 .col_int_nokey ) AS SQ1_field1
FROM ( t2 AS SQ1_alias1 RIGHT JOIN t1 AS SQ1_alias2
ON ( SQ1_alias2.col_varchar_key = SQ1_alias1.col_varchar_nokey )
)
WHERE SQ1_alias2.pk < alias1.col_int_nokey OR alias1.pk
) AS field1
FROM ( t1 AS alias1 JOIN t2 AS alias2 ON alias2.pk )
GROUP BY field1
';

EXECUTE stmt;
EXECUTE stmt;

DEALLOCATE PREPARE stmt;

DROP TABLE t1, t2;

If to replace the conjunctive degenerated condition alias1.pk for the equivalent predicate alias.pk<>0
then there is no problem with the second execution of the PS for the query:

MariaDB [test]> PREPARE stmt FROM '
'> SELECT
'> ( SELECT MAX( SQ1_alias2 .col_int_nokey ) AS SQ1_field1
'> FROM ( t2 AS SQ1_alias1 RIGHT JOIN t1 AS SQ1_alias2
'> ON ( SQ1_alias2.col_varchar_key = SQ1_alias1.col_varchar_nokey )
'> )
'> WHERE SQ1_alias2.pk < alias1.col_int_nokey OR alias1.pk<>0
'> ) AS field1
'> FROM ( t1 AS alias1 JOIN t2 AS alias2 ON alias2.pk )
'> GROUP BY field1
'> ';
Query OK, 0 rows affected (0.01 sec)
Statement prepared

MariaDB [test]> EXECUTE stmt;
--------

--------

--------
1 row in set (0.00 sec)

MariaDB [test]> EXECUTE stmt;
--------

--------

--------
1 row in set (0.01 sec)

(See also bug #12582849 from the mysql code line)