DSN of the IBM loadlib which houses the SVC modules

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

What is the DSN of the IBM loadlib which houses the SVC modules?

Are z/OS 1.7 SVCs able to run for Amode31 programs which invoke them from above 16MB?

Thanks.

dick scherrer · Posted: Fri Aug 29, 2008 1:38 am

Hello,

1. Do you know the name of one of the modules you are looking for? Have you tried ISRDDN? What are you actually looking for - svc's live multiple places.

2. You might check with your system support people and/or ibm.

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

Module name could be IEASVCxx.
ISRDDN only works if the PDS is already allocated within the ISPF session.
If you say SVC routines are stored in multiple PDS'es, how about suggest one?

dick scherrer · Posted: Fri Aug 29, 2008 6:01 am

Hello,

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

Thx. Too many intermediate persons would be involved to contact IBM. I did a QW against "SVC" and all IBM gave was a list of functions agaginst a list of SVC numbers.

In the old days of DOS/VSE IBM manuals would be quoting IG*** module names at the drop of a hat. Can you think of an IBM manual name which would state whether SVCs imbedded inside application programs running under Amode31 Rmode31 z/OS 1.7 would work just as well above 16MB as below?

dick scherrer · Posted: Fri Aug 29, 2008 7:44 am

Hello,

You might try here:
publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/IEA2V270/1.4.2.2?

If you follow the "open" svc, it goes here (which talks about mode=24 or 31):
publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/IEA2V270/1.4.3.20?DT=20060606170900#HDRSVC19

There are links to other publications as well throughout.

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

Thanks Dick,

I shall chase the manuals in that neck of the woods. The right IBM manual is not always easy to pinpoint when search is by keyword and a large number of manuals are returned.

Looks like z/OS SVC is builtin ambidextrous, able to work with both Amode31 and Amode24. I need to ascertain this is so for a total of 14 SVCs, so a categorical IBM statement that SVCs are now ambidextrous is what I shall need.

dick scherrer · Posted: Fri Aug 29, 2008 8:45 pm

You're welcome - good luck

Just a thought - if this were a problem, search engine hits would probably be high. . .

d

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

LOL, probably true, but as the hanging judge said,

"absence of proof is not proof of absence."

nevilh · Active User Joined: 01 Sep 2006 Posts: 262

SVC location depends on how they are defined in member IEASVCxx in PARMLIB. Type 2 and type 6 svc's are found in NUCLEUS. They are named SVCxxx where xxx is the svc number. Other svc's are normally found in LPALIB concatenation. These should follow the following naming conventions IGC00xxy where xx is the first 2 digits of the svc number and y is one of the following (abcdefghi or one other character x'C0') So svc 231 would be IGC0023A svc 232 would be IGC0023B . Module names can be specified in the IEASVCxx that override these defaults.
On top of this you also have extended svc's normally named IGX000xx (lpalib concatenation) where xx is any value between 00 an 99. Then there are the people who like to load the svc's dynamically from a program then the svc's can reside in any authorised library. SVC's should work whatever amode/rmode the invoking program has.

nevilh · Active User Joined: 01 Sep 2006 Posts: 262

Mistermind · New User Joined: 08 Feb 2008 Posts: 46 Location: Dublin

Thanks for the info. When back at the office I shall look at the module amode/rmode attributes from the appropriate loadlib directories, then issue resolved..

Slightly surprised that Third Party Software has been admitted into IBM SVCs. Hope they are up-to-date in step with IBM practice.