String concat issues in 8.3.1 (8.3.1 bug?)

I have a table as follows:

\d test_dim
Table "public.test_dim"
Column | Type | Modifiers
-----------------+-----------------------------+------------------------
customer_srcid | bigint | not null
segment_srcid | bigint | not null
show_name | character varying(500) | not null
create_dt | timestamp without time zone | not null default now()

I have this table in a db on an 8.2.6 cluster and in another db on an
8.3.1 cluster.

In both clusters I can do the following:

# select customer_srcid, segment_srcid from test_dim;
customer_srcid | segment_srcid
----------------+---------------
1 | 1
1 | 2
1 | 3
1 |
1 |
1 |
1 | 1
1 | 2
1 | 3
1 |
1 |
1 |
1 | 1
1 | 2
1 | 3
1 |
1 |
1 |
1 | 1
1 | 2
1 | 3
1 |
1 |
1 |
(24 rows)

In version 8.2.6 I can do this:

# select customer_srcid || segment_srcid from test_dim;
?column?
----------
11
12
13

11
12
13

11
12
13

11
12
13

(24 rows)

However in the 8.3.1 cluster I get this (Both clusters have the same
data set - via a pg_dump):

# select customer_srcid || segment_srcid from test_dim;
ERROR: operator does not exist: bigint || bigint
LINE 1: select customer_srcid || segment_srcid from sl_segment_dim;
^
HINT: No operator matches the given name and argument type(s). You
might need to add explicit type casts.

Thoughts ?

Thanks in advance

/Kevin