The only way I have seen original abends is to go into the DSA or CAA control block in IPCS to see what really happened.
....U4038 A severe (unhandled) error occurred, but A no dump was requested (useless) ....U4039 A severe error occurred and a CEEDUMP A (and optionally System dump) was requested ....U4083 Save area backchain in error ....U4087 Error during error processing ....U4093 Error during initialization ....U4094 Error during termination Lizette > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Blaicher, Christopher Y. > Sent: Monday, November 03, 2014 8:31 AM > To: [email protected] > Subject: Re: Abend s0077 > > Peter, > We are almost exclusively an assembler shop, but recently we have added a few C > routines that use LE. > > It blows me away that LE has to take a perfectly good 0C1, 0C4 or 0C7 and convert > it into a U4xxx code. Not only that, they have to obfuscate the registers. > > Is there a conversion table for LE user codes to regular abend codes? (OK, 0Cx's > aren't an abend in that no ABEND macro was issued, but for 50 years we have > been calling them ABENDS0C4 and the like.) > > Chris Blaicher > Principal Software Engineer, Software Development Syncsort Incorporated > 50 Tice Boulevard, Woodcliff Lake, NJ 07677 > P: 201-930-8260 | M: 512-627-3803 > E: [email protected] > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Peter Relson > Sent: Monday, November 03, 2014 8:58 AM > To: [email protected] > Subject: Re: Abend s0077 > > >But I AM the frequently the first person to look it up in the manual. > > I understand that in your shop you are the first to look at the book. But that is not the > same as being the first to see the abend. Even on IBM-Main there are many cases > of ragging on people to RT(F)M (at least to take a stab at finding what to read). If > any of your customers are writing and running their own programs, that should be > expected. If they are just running canned jobs that "should" never abend (or if they > do, then it's not the customer's fault) then passing it along to the sysprog makes > perfect sense. The books have no way of knowing whether they are being read by > someone who wrote the program that blew up or someone who just ran someone > else's program which blew up, so the books take a rational approach at segregating > the information according to the potential audiences. > > >hiding an S0C7 as a U4xxx error is not helpful. > > Are you suggesting something like when LE recovery fields a system-produced > completion code (S0C7 is not an "abend") that it leave that code alone and not > change it to a user completion code? That could be provided as some sort of a > configuration option. It cannot be done unconditionally as it is compatible and could > break existing programs. It might not even be possible if you are using LE (E)SPIE > since there is no > 0C7 in such a case, there is just a program interrupt 7 presented to the ESPIE > routine (but as of a few releases ago, LE could for such a case tell the system to > continue on to RTM for this program interrupt where it would become S0C7). If > "leave it alone" is something that you want, then I suggest that you go through a > more formal approach to request it than an IBM-Main conversation. > > Peter Relson > z/OS Core Technology Design > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
