All I can suggest is the following:
- Run an AMBLIST to find out how many CSECT names there are in each of
your LMODS.
- Disassemble each CSECT in turn and then try to figure out what Fortran
source code generated it.
I did come up against a similar problem more than 20 years ago, caused
by a loop in some Fortran code from a 3rd party ISV and executing in its
own address space. The ISV could not identify what caused the loop (it
happened at random) and would not release the source code either. I
tracked down the start/end offsets of the loop in the LMOD from a full
system trace, called in the ISV and asked them to recompile their source
with an assembly listing (compiler 'LIST' option?) From this I could
then match the object code in the dump with the Fortran assembly
listing, and thus fix the problem.
You could also try XDC'ing your LMODs if that helps.
Just my ha'pennyworth.
Cheers, Chris Poncelet
Itschak Mugzach wrote:
I
have three (quit complex) FORTRAN load modules which I do not have the
source code of. I Know the Source company, but they deal with Cobol, and I
need Fortran.
Any idea how to reverse engineering the load modules?
ITschak
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
