RE: ILE Debugger and OPM Programs

Hi Rory

Thanks, I guessed the jumping back to the PLIST was normal - a bit like
jumping back to the File specs when doing I/O and the *NODEBUGIO option
isn't set (that really winds me up)

The majority of the application is in 2E generated RPG/400 although there
are the more recent functions generated in RPG IV - I daren't say ILE as
very little of the system (even the "external" programs use real ILE
techniques. We actually had a directive last year not to use RPG IV or ILE
as too many of the programmers were unfamiliar with it. Unbelievable.

All the best

Jonathan


-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-bounces@xxxxxxxxxxxx]
On Behalf Of Rory Hewitt
Sent: 01 March 2011 16:30
To: RPG programming on the IBM i / System i
Subject: Re: ILE Debugger and OPM Programs

Jonathan,

I've just been debugging 2E programs too :)

The behavior where the debugger jumps to the *ENTRY PLIST line and then back
(and/or sometimes to the last line of the source - probably the ZZINIT
subroutine ENDSR line) is 'normal' behavior with OPM programs when using the
source debugger.

One quick thought - if the program has a parameter list (which it will), the
source debugger will seem to 'hang' for each of the parameters. In other
words, you press F10 in the caller on the line before the CALL and the CALL
line is highlighted. If the called program has 5 parameters, you need to
press F10 5 times before the program is actually called and the line after
the CALL statement is reached. This is also affected by whether the
parameters are input, output or both. In other words, for a call to a
program passing 5 parameters, you may need to press F10 4 times and F22 on
the 5th time. Or something like that.

That's why I always explicitly add each OPM program to debug.

Rory

p.s. Why not just generate RPGIV source from 2E? It's been around for many
years now, so the bugs have been worked out.

From: Jonathan Mason

To:
Cc:
Sent: Tue, 01 Mar 2011 15:13
Subject: ILE Debugger and OPM Programs

Hi All

A colleague of mine is having an issue trying to debug an OPM program
using the ILE Debugger and it has us both completely stumped.

He has set the debugger to start debugging at the first program in his
process and is then using F22 to step into each program as he reaches
them. For the second program this worked fine, but for the third
program the debugger just jumps straight to the next line of the current
program rather than stepping into program three.

All of the programs are OPM (generated by Synon/2e) and he is using the
STRDBG command with OPMSRC(*YES) specified. Each of the programs has
been recompiled into a testing library with the *SRCDBG option set so
the debugger should pick it up.

In fact if he presses F14 within the source display and adds the third
program to the list of programs he can display the source and add
breakpoints. However the source doesn't get entered into.

The call to program three does carry an error indicator (*IN90) which is
off when the next line is reached, also the default Synon return code
for the called program is blank indicating successful completion.

The correct version of Program Three is at the top of the library list,
but we can't see any reason why the debugger would appear to step over
the program call. When F10 (or F22) is pressed to step into the third
program the display does jump back to the *ENTRY parameter list of the
second program before another F10 moves it back to the line after the
call.

Any help would be gratefully appreciated.

All the best

Jonathan