The problem is the message, not the ABEND code. And, Yes, the message after a program check (ABEND) should reflect the original interrupt code (ABEND code and reason code).
Unless the user turned off SPIE, the user should never see an ABEND S0Cx intercepted by LE; and, yes, the LE message for a program check intercepted by LE should include more than the interrupt code. Memory is a lot less expensive than it was in the S/360 days, and this is an issue that goes well beyond LE. Has anybody submitted an RFE asking that messages include more than just the numeric codes and variable data? Enhancing the message text would certainly make the platform more accessible to newbies. -- Shmuel (Seymour J.) Metz http://mason.gmu.edu/~smetz3 ________________________________________ From: IBM Mainframe Discussion List [[email protected]] on behalf of Clark Morris [[email protected]] Sent: Monday, June 7, 2021 9:05 PM To: [email protected] Subject: Re: U 4087 abend [Default] On 7 Jun 2021 15:25:53 -0700, in bit.listserv.ibm-main [email protected] (Seymour J Metz) wrote: >After all, the compilers, run time environments and utilities haven't been >using user ABEND codes for much more than half a century. > What I object to is LE obscuring the original abend code with an abend code that doesn't have a 1 to 1 relation to the original abend. A U4xxx-S0C7 message would be adequate. In regard to the application programmer in the original posting, Has he or she looked at all of the documentation such as CEEDUMP which as I recall has the original abend code? Is that programmer knowledgeable enough to know that a S0C7 is an application error 99.999 percent of the time (there may be a rare instance where it isn't). Clark Morris > >-- >Shmuel (Seymour J.) Metz >http://mason.gmu.edu/~smetz3 > >________________________________________ >From: IBM Mainframe Discussion List [[email protected]] on behalf of >Charles Mills [[email protected]] >Sent: Monday, June 7, 2021 5:08 PM >To: [email protected] >Subject: Re: U 4087 abend > >Well, the LE folks did not do anything for the clarity of this issue. I would >call an LE failure a system ABEND, but they are all Unnnn ABENDs. > >Charles > > >-----Original Message----- >From: IBM Mainframe Discussion List [mailto:[email protected]] On >Behalf Of zMan >Sent: Monday, June 7, 2021 10:56 AM >To: [email protected] >Subject: Re: U 4087 abend > >Well, for my money, any ABEND starting with U *must *be a system ABEND. >Those ones starting with S are clearly something else. > >On Mon, Jun 7, 2021 at 12:36 PM Seymour J Metz <[email protected]> wrote: > >> Did the developer read the message description and look at the dump? Did >> he give a cogent reason for believing it was a systems error? If not, then >> the smart money says that it's his error. >> >> "Who knows, the horse might learn to sing." >> >> I would look for storage overlay, but that's not the only thing that >> could cause recurse condition handling. >> >> >> -- >> Shmuel (Seymour J.) Metz >> http://mason.gmu.edu/~smetz3 >> >> ________________________________________ >> From: IBM Mainframe Discussion List [[email protected]] on behalf >> of Bill Giannelli [[email protected]] >> Sent: Monday, June 7, 2021 11:41 AM >> To: [email protected] >> Subject: U 4087 abend >> >> we have a job that keeps abending in an application program that the >> developer insists is a "system" issue. what does a U 4087 abend indicate? >> thanks >> Bill >> >> ---------------------------------------------------------------------- >> For IBM-MAIN subscribe / signoff / archive access instructions, >> send email to [email protected] with the message: INFO IBM-MAIN >> >> ---------------------------------------------------------------------- >> For IBM-MAIN subscribe / signoff / archive access instructions, >> send email to [email protected] with the message: INFO IBM-MAIN >> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
