Hey Ed, thanks for the quick answer.

Unfortunately NI SDWAERRD,X'FF'-SDWACLUP did not seem to make any
difference.

Perhaps I am wrong in my whole model of what is going on. I can't see the
higher ESTAE(X)(s) if any; I am just inferring their existence and
parameters. If I don't do any of this, if I don't issue any ESTAE(X), then
*something* gets control on "normal" ABENDs but apparently not on operator
cancels. (The code is LE with TRAP(ON,SPIE). LE is documented as not
processing Sx22 ABENDs.)

Incidentally, setting a retry point and issuing a new ABEND there did not
work either, contrary to my hopes.

BTW, the issue is not Sx22 abend codes per se. If I issue ABEND
X'122',,,SYSTEM (from my code, but outside of ESTAE(X) recovery) then LE
condition handling intercepts it.

Charles

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On
Behalf Of Ed Jaffe
Sent: Tuesday, August 27, 2013 4:56 PM
To: [email protected]
Subject: Re: Questions about ESTAE(X)

On 8/27/2013 4:52 PM, Charles Mills wrote:
> Question 1: Just want to confirm that I am reading the documentation 
> correctly On ESTAE(X), TERM=YES basically says "Yes, I want my 
> recovery routine driven for operator cancel and things like that" and 
> TERM=NO says "only drive the recovery routine for 'normal' ABENDs." Is 
> that basically correct?

Yes.

> Question 2: (Assuming I am reading TERM=YES correctly) if an ESTAE(X) 
> TERM=YES is chained after an ESTAE(X) TERM=NO, is there any way to get 
> the chained recovery routine to percolate a TERM=YES-type ABEND?

My understanding is that a TERM=NO exit is skipped when SDWACLUP=ON. 
But, that understanding comes strictly from empirical evidence. I have not
seen it in the code...

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to