RE: source line # in a dump for an SQLRPGLE pgm

["rich herdman" <rherdman@xxxxxxxxxxxxxxx>]

That's what I did.  Not having the compile listing from two years ago, I
recompiled it (since *NOGEN doesn't work in SQLRPGLE, I compile it
to QTEMP) and used that compile listing to look for the statement
in error reffered to in the dump.  That compile listing says that the line #
I am looking for is a READE, which should not generate a 'target too small
for numeric operation'.

Since the object was created under V4R3 and my compile listing that I am using
today was generated under V5R1, I wonder if the differences between version is
causing the inaccurate line #'s.



On Mon, 18 Mar 2002 10:45:16 -0500, Buck Calabro wrote:

>Rich Herdman wrote:
>
>>Anyways, I dumped the program to get the
>>statement # that it blew up on.  The
>>problem is that I don't think the
>>line number is right.
>
>The statement in the error message and dump matches the compiler listing
>unless you specify
>h option(*srcstmt).  Also be aware that the line number in messages may not
>be accurate if you have an optimised program.
>
>If you don't have a compiler listing on file, re-compile the program into
>QTEMP to get a new listing without creating a new object.  If it were a
>straight RPG program you could use OPTION(*NOGEN), but for an SQL compile
>*NOGEN just runs the preprocessor.
>  --buck


Rich Herdman
email: rherdman@columbus.rr.com
www:  http://home.columbus.rr.com/herdman/

"It never does to leave a live Dragon out
of the equation." - Bilbo Baggins, 'The Hobbit'