Ron and Jenny Hawkins wrote:
Jim,
Thanks for that - so it is more like a bus than a train :-)
Is there any frequency or time associated with the period between PCIs?
In the seldom-ending channel programs I've seen, the PCI is inserted at
the point where the end of the CCW chain would normally occur for a
traditional "short-lived" I/O. When handling the interrupt, the
reads/writes just performed are posted complete and pending requests, in
the form of more CCWs, are inserted into the chain. (The insertion need
not be at the end of the chain). This processing is precisely what makes
the I/O seldom-ending. The benefit is considerably less IOS overhead per
request.
FWIW, this is the technique used by JES3 for all spool I/O. On a busy
system, a single STARTIO macro can run for minutes or even hours.
Pending requests are inserted into the already-running I/O based on
current disk position. This optimization made a tremendous difference on
SLED disks.
--
.-----------------------------------------------------------------.
| Edward E. Jaffe | |
| Mgr, Research & Development | [EMAIL PROTECTED] |
| Phoenix Software International | Tel: (310) 338-0400 x318 |
| 5200 W Century Blvd, Suite 800 | Fax: (310) 338-0801 |
| Los Angeles, CA 90045 | http://www.phoenixsoftware.com |
'-----------------------------------------------------------------'
----------------------------------------------------------------------
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
