CRTRPGMOD Compiler issue

CRTRPGMOD runs for one particular source member (whether it's batch or 
interactive) and "..never.." finishes!  It apparently is getting just stuck. 
The very last line of source prints, copy books included, nothing recursive, 
and then just hangs there at the same statements of IBM's compiler programs: 

Library QDEVTOOLS program QRN4CPP, procedure "main", statement 519, and 
Library QRPGLE program QRN4MAIN, procedure "main" at statement 140, procedure 
"dicmerg" at 25.

Has anyone else come up with this? Is there maybe a specific PTF to this 
specific problem? We're getting the latest cum-ptf's in a few weeks....

The command parameters are all "standard":

CRTRPGMOD 
        MODULE(Lib/Mod001) 
        SRCFILE(SourceLib/SourceFile) 
        SRCMBR(SourceMbr)
        OPTION(*NODEBUGIO)
        DBGVIEW(*LIST)
        TGTRLS(*CURRENT)

Changing to DBGVIEW(*SOURCE)  and then DBGVIEW(*NONE) did not help.

The "sys-admin" guy here says we're getting the latest cum's in a few weeks. 
I'm going to find a way around this than spend time with IBM, for timelines and 
other reasons. Nobody else encountered this here, and my other code compiles 
like a charm, so it must be unique to the code in this particular source.

My strategy right now is to (1) rearrange the code and compile, and it that's 
not successful, (2) just rewrite it and see what happens. 

Once upon a time I had a program doing funny things with field values as debug 
showed, and simply moving one of the affected data structures to another place 
in the D-specs resolved the situation...

--Alan