Weird SQL results

I am running a big SQL that is giving me bogus results. I have narrowed it
down to this chunk. I have hardcoded some values because they come from
other parts of the SQL. The group by is because there are some fields
summed that I dropped from the select to keep it minimal.



with bmi_info as
(
select program_name
bound_module_library,
bound_module,
source_file,
source_file_member
from bound_module_info
where program_library = 'IPLIBP'
group by program_name,
bound_module_library,
bound_module,
source_file,
source_file_member
)
select bmi.*
from table(object_statistics('IPLIBP','*MODULE','AEEC10BR')) stat
join bmi_info bmi
on bmi.bound_module = stat.objname;



If I run it with the bound_module_library variable, I get one record.

Bound Module Source File Source File Member

AEEC10BR QRPGLESRC AEEC10BR



If I run it without the bound_module_library variable, I get two records.

Bound Module Source File Source File Member

AEEC10BR QRPGLESRC AEEC10BR

AEEC10BR QRPGLESRC AEEN01AR <-- No idea where
this row is coming from





Why would adding the extra variable to the select and group by REDUCE the
number of records? If anything, I would expect the opposite if there were
multiple libraries.



Also, I ran this sql to look for the bogus record and it returned an empty
result set.

select *
from bound_module_info
where source_file_member = 'AEEN01AR'
and bound_module = 'AEEC10BR'



Is there something I am overlooking (or doing stupid) or does our database
have issues?