I think it is worth pointing out the most likely cleaning up logic used for 
enqueues (note that I do not have access to GRS source - but this is how I am 
guessing things work).

GRS will establish a TYPE=ADDRSPC and a TYPE=TASK resource manager routine to 
cover all address spaces and all tasks in the system.

These two resource manager routines give GRS a fighting chance to maintain 
integrity when "ungraceful" termination occurs for either the task (TCB) that 
requested an enqueue or the entire address space of the requestor.

When the task terminates, the task level resource manager gets control in the 
home address space and (as long as there is enough wiggle room in the private 
area) it can send a "clean-up resources for this TCB in this ASID" request over 
to GRS (most likely via some sort of PC-ss).
This is good news as it not only informs GRS of tasks that are abending - but 
also catches the situation during normal termination when the task did not 
issue a dequeue for all enqueues previously obtained.
These resource managers are protected from cancels, detaches and any "SRB to 
task" percolated abends - in other words they are pretty much guaranteed to run.

If the ASID terminates, the ASID level resource manager gets control in the 
*MASTER* address space (ASID 0001) and it can send a "clean-up resources for 
the ASID" request over to GRS (again most likely by a PC-ss).
Another piece of good news as a belt-and-braces approach to GRS structure 
integrity is beneficial to long term system up-time.

No matter what you fiddle about with in SWA, these GRS resource manager 
routines will enable GRS to cleanup any underlying resources obtained by the 
task or address space.

Note that unlike enqueues, GRS latches do NOT get automatically cleaned up 
during task termination.



Rob Scott
Principal Software Engineer
Rocket Software
77 Fourth Avenue . Suite 100 . Waltham . MA 02451-1468 . USA
Tel: +1.781.684.2305
Email: [email protected]
Web: www.rocketsoftware.com


-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf 
Of Steff Gladstone
Sent: 21 April 2015 16:12
To: [email protected]
Subject: Re: ENQ for the life of the job

Sorry, John. My question was unclear.  It was referring to your earlier remarks 
above:

    Thinking about it a bit more, given what Mr. Relson
     said about RTM, doing this _should_ work even if the initiator terminates
     abnormally since RTM should clean up the ENQs during EOM processing.

Are you saying that RTM should clean up the ENQ even without modifying the SWA 
with information for the DD, as you suggested to do earlier?

On Tue, Apr 21, 2015 at 4:42 PM, John McKown <[email protected]>
wrote:

> On Tue, Apr 21, 2015 at 7:00 AM, Steff Gladstone <
> [email protected]>
> wrote:
>
> > Even without modifying the SWA with information for the DD, as you
> > suggested earlier?
> >
>
> ​Yes, you can do the directed ENQ yourself without modifying the SWA.
> Which, by the by, I _mentioned_, not _suggested_. At least, I didn't
> mean for it to be a suggestions. It's the fly with a shotgun ​
> approach. ​The main problem with the directed ENQ to the initiator TCB
> is still "how to get the DEQ done at end of job?". That is the "sticking"
> point in this whole thing.
>
> --
> If you sent twitter messages while exploring, are you on a textpedition?
>
> He's about as useful as a wax frying pan.
>
> 10 to the 12th power microphones = 1 Megaphone
>
> Maranatha! <><
> John McKown
>
> ----------------------------------------------------------------------
> 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
================================
Rocket Software, Inc. and subsidiaries ■ 77 Fourth Avenue, Waltham MA 02451 ■ 
+1 800.966.3270 ■ +1 781.577.4321
Unsubscribe From Commercial Email – [email protected]
Manage Your Subscription Preferences - 
http://info.rocketsoftware.com/GlobalSubscriptionManagementEmailFooter_SubscriptionCenter.html
Privacy Policy - http://www.rocketsoftware.com/company/legal/privacy-policy
================================

This communication and any attachments may contain confidential information of 
Rocket Software, Inc. All unauthorized use, disclosure or distribution is 
prohibited. If you are not the intended recipient, please notify Rocket 
Software immediately and destroy all copies of this communication. Thank you.

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

Reply via email to