While not presuming to understand even 10% of what the Piper writes, I'm
curious what level of Pipes you are running?
The PIPELINES NEWS file containing in its disclaimer the text:
---<snip>---
The "CMS Pipelines Run-time Library", which currently is CMS Pipelines
sublevel 07 as of November 1, 2010, is available for download from the
CMS Pipelines home page.
---<snip>---
IIRC that PIPELINE NEWS file, matches:
PIPINX560I CMS/TSO Pipelines, 5654-030/5655-A17 level 110C0008.
PIPINX086I CMS/TSO Pipelines, 5654-030/5655-A17 1.0112 (Version.Release/Mod) -
Generated 3 Dec 2010 at 11:10:08
Source: Princeton Runtime Distribution
And somewhat more detailed WAITDEV documentation of:
---<snip>---
WAITDEV <device>
Wait for an interrupt on the specified device.
WAITDEV shorts its input to its output and terminates
immediately if a real device is associated with the virtual
device.
Otherwise it waits for the device and writes the 12-byte SCSB
to the output when an interrupt is received.
The three cases can be distinguished if a null record is
supplied on input.
Valid virtual devices are terminals, graphs, readers, and
CTCAs. You'd typically guard a WAITDEV with some kind of
timeout. Otherwise PIPMOD STOP will do the job.
Note that coupling a CTCA does not generate an interrupt (but
writing from the other side generates attention).
---<snip>---
Mike Walter
Aon Corporation
The opinions expressed herein are mine alone, not my employer's.
-----Original Message-----
From: CMSTSO Pipelines Discussion List [mailto:[email protected]] On
Behalf Of Paul Gilmartin
Sent: Monday, August 08, 2011 9:46 AM
To: [email protected]
Subject: Re: Reader Interrupts
On Aug 8, 2011, at 00:34, John P. Hartmann wrote:
> WAIDEV
>
(Or, perhaps, WAITDEV.)
> See pipeline news.
>
Thanks. AHELP tells me:
...
! 1. A real device is already present for the virtual device. This means that
! a terminal is already present. This cannot occur for readers and channel
! to channel adapters.
Is this because those device types are obsolete? Might this not
yet happen in a second level VM if a virtual reader defined at
first level is ATTACHed to the second level VM, where it would
appear real? Or is support for real readers completely absent
from CP nowadays?
> On 8 August 2011 03:33, Glenn Knickerbocker <[email protected]> wrote:
>> On Sun, 7 Aug 2011 14:46:32 -0600, gil wrote:
>>
>> Rather than reinvent the whole server, check out PIPESERV:
>>
>> http://www.vm.ibm.com/download/packages/descript.cgi?PIPESERV
>>
Thanks also. I need to check which is more suitable.
A broader question. Does WAITDEV (and perhaps other builtin
stages) modify the I/O new PSW? Do they restore it before
running CMS or CMD? (CP shouldn't matter.) Is there a risk
that interrupts will be lost while CMS owns the PSW? Is
there a compatibility matrix of Pipeline stages versus CMS
commands available?
Thanks,
gil
