> I am trying to set a SLIP trap for a SRB routine (APXABBVS) that runs in > the JES2 address space. The trap is not intercepting the abend. The > SYSUDUMP shows the S0C4 location after the SCHEDULE macro in the main > program. Any thoughts? > > SLIP SET,C=0C4,ID=MMS2,LPAMOD=APXABBVS,J=JES2,ACTION=SVCD,END
First; (editorial comment) you probably should not have ANY dump DDs in your long running address spaces. They are about as useful as an ashtray on a motorbike. The abend that is being "seen" here is the result of SRB to task percolation and actually has nothing at all to do with the location after the SCHEDULE macro. So we can tell that an SRB *did* get a S0C4 somewhere in that address space and that the SLIP trap did not match. The more specific you are (e.g. by specifying LPAMOD) the less likely you are to get a SLIP trap to hit, e.g. the one above would only fire if the abend PSW was actually within that module. If it had called other code or system services and the failure occurred anywhere outside of that LPA module you won't see it at all. Rather than guess, or fool about trying to get address ranges etc. I would just remove the LPAMOD keyword and let it snag a dump for the first S0C4 in an SRB by adding "MODE=SRB". You will also want to check your default SVC dump options to make sure you get the data you need, or tailor it via SDATA keyword of the SLIP cmd. Once you have a dump, take a look where it occurred and see if that's the problem you're interested in. If it is you're done. Otherwise you can start to narrow in on where it is really happening and add some RANGE keywords. Beware that there can be many abends that match, so you may have to wash/rinse/repeat a number of times to figure out. CC
