Path: csiph.com!1.us.feeder.erje.net!feeder.erje.net!1.eu.feeder.erje.net!newsfeed.fsmpi.rwth-aachen.de!newsfeed.straub-nv.de!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail From: Anne & Lynn Wheeler Newsgroups: alt.folklore.computers,comp.sys.raspberry-pi Subject: Re: Is it a lost cause? Date: Sat, 02 Jul 2016 08:13:54 -0700 Organization: Wheeler&Wheeler Lines: 33 Message-ID: <87twg8xebx.fsf@garlic.com> References: <20160613210220.7d9874b5@maxa-pc.cpe.bvcom.net> Mime-Version: 1.0 Content-Type: text/plain Injection-Info: mx02.eternal-september.org; posting-host="6a701f7a2f120c3c3ec794af16136f6c"; logging-data="18215"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX19L4QPeySHw9G7aQusNwDdsZYUY1yAmvwc=" User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux) Cancel-Lock: sha1:PRJ7XemfuEhTA7G1mO9xHwqWkmk= sha1:EvVmdkcEKG2Kl21Ff/bDWfiGmp8= Xref: csiph.com alt.folklore.computers:166311 comp.sys.raspberry-pi:12271 Dennis Lee Bieber writes: > Sigmas were (in)famous for having something like 224 interrupt vectors > (when IBM 360/370 only had 7 and had to poll to determine which device > interrupted). Each interrupt vector basically held a program status > doubleword, so an exchange PSD made for fast response. 360/370 had svc, i/o, program, external/timer, machine-check interrupts ... loaded new psw (instruction address) from low store at same time as storing old psw (interrupted instruction address). interrupts stored information about the interrupt, in the case of i/o interrupt, it stored the interrupting device address (channel, controller, device) and the status (channel status word). old ios3270 version of "green card" that I did q&d convert to html low store layout for interrupting, old/new psw, csw, etc http://www.garlic.com/~lynn/gcard.html#4 channel status word (for i/o interrupt) http://www.garlic.com/~lynn/gcard.html#6 certain kinds of remote terminals had several termainals connected to single device address/port/line ... so there was additional work to find out which of several remote terminals ... sharing same line ... was responsible. note that 360/370 interrupts didn't save/restore registers ... so there was standard interrupt handlers to store registers of the interrupted process and load registers for the interrupt handler. -- virtualization experience starting Jan1968, online at home since Mar1970