Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83 (MC830713); site vu44.UUCP
Path: utzoo!linus!decvax!ittvax!dcdwest!sdcsvax!sdcrdcf!hplabs!hao!seismo!mcvax!vu44!jack
From: jack@vu44.UUCP (Jack Jansen)
Newsgroups: net.micro.16k
Subject: 16202 ICU problems.
Message-ID: <412@vu44.UUCP>
Date: Wed, 3-Oct-84 08:52:56 EDT
Article-I.D.: vu44.412
Posted: Wed Oct  3 08:52:56 1984
Date-Received: Fri, 5-Oct-84 20:22:57 EDT
Organization: VU Informatica, Amsterdam
Lines: 21

We are currently designing a 16k multiprocessor system, and one
of the problems we are facing is the following :
According to the manual, the 16202 ICU *always* latches
interrupt requests, even if you specify interrupts as
level-triggered. This is a problem to us, since it makes it very
difficult to make sure that an interrupt is serviced by exactly
one CPU.
Question: Do I read the manual correctly? Is there any way to
remove an interrupt request from an ICU, after you have offered
it to him (but before the CPU has started an interrupt cycle)?
The other solution is to keep the ICU from seeing the interrupt
request if it's CPU is not willing to service interrupts, but
how can I see that from the outside???

If anyone has hints, solutions, rumours, flames or whatever I'll
be very glad to recieve them. If there are other people with the
same problem, drop me a line and I'll keep you up to date to the
replies I get.

	Jack Jansen, {seismo|philabs|decvax}!mcvax!vu44!jack
	or				       ...!vu44!htsa!jack