On Tue, 5 Jan 2010 00:15:44 -0600, Angel-Luis Dominguez
<[email protected]> wrote:

>I am developping ICSF service exits to write information in SMF.
>
>Manuals along all the releases only say:
>
>"The installation exit gets passed the address of the service parameter list
>in Register 1."
>
>But I have encountered that ......
>
>If you have HCR7740 function, the parameter list addressed by R1 is in
>the ASID of ICSF.
>
>But if you migrate to HCR7750 function, the exits that are working fine
>starts to have S0C4 abend's and others, even you have ptf's arising this
>applied..
>
>That is due to the list pointed by R1 isn't  in ICSF ASID and is now in the
>ASID of the user requiring ICSF services.
>
>I haven't found any documentation of this change and I needed serveral
>days to detect the change thinking it was a problem of my code instead a
>change in funcionality.
>
>Apar OA23636 points to this but was applied.  The key for resolving was
>in a comment in apar OA23955. Thats it's the text of the comment:
>
>"I have added a " LAM AR1,AR1,=F'1' " instruction near the start of the
>User Service Routine and this has corrected the abend"

The book for z/OS V1 R11 (ICSF FMID HCR7770, see
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/CSFB2ZA0/5.3.2?SHELF=EZ2ZBK0H&DT=20091114130346
or http://preview.tinyurl.com/yeb9hnf ) and as I read it the book clearly
states that the service exits run in cross-memory mode and must specify an
ALET of 1 to access the parameters, which will be in the secondary address
space.  That same text is in the ICSF books at least as far back as z/OS V1 R7.

APAR OA23636 addresses an error in Omegamon/XE where the exits it provides
were not properly using the ALET of 1, and via that APAR they corrected
their exits to provide that ALET.

-- 
Walt 

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to