In article <[email protected]> you wrote: > Den fredag 11 september 2015 kl. 14:40:55 UTC+2 skrev Michael Knigge: > > All, > > > > I?ve reimplemented the SWAREQ-Macro in Java (my ?template? was the well > > known REXX from Gilbert Saint-Flour, see > > https://www.mail-archive.com/[email protected]/msg30741.html). > > > > Since z/OS 2.2 this implementation works no longer. The address to the QMAT > > can be a 64 bit address since z/OS 2.2. No problem so far. But it seems to > > me the complete logic > > > > DO WHILE sva>65536 > > qmat = C2D(STORAGE(D2X(qmat+12),4)) > > sva=sva-65536 > > END > > RETURN C2D(STORAGE(D2X(qmat+sva+1),4))+16 > > > > seems to work no longer?. I guess the problem is that the ?return value? > > from SWAREQ may be a 64 bit address but the ?old? routine only gets a 32 > > bit value at qmat+sva+1. ?
> Hello > If you have jobclass SWA ABOVE in z/OS 2.2 old stuff like CBT pgm CONCATEM > will abend with 0C4. I have not seen any "warning" from IBM about the fact > that ABOVE means that SWA is placed above the 2GB bar and how it will affect > user written programs. The closest sign is APAR OA48915. > Workaround is of course to keep JOBCLASS SWA BELOW on those who fails > Best Regards > Bosse Stamhagen You're responding to an old post. I don't know about CONCATEM, but I fixed SWAREQ last October. It's on the current CBT183 file as SWAREQ22 (but it should also work for pre-2.2 systems). -- Don Poitras - SAS Development - SAS Institute Inc. - SAS Campus Drive [email protected] (919) 531-5637 Cary, NC 27513 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
