Return to BSD News archive
Xref: sserve comp.periphs:4128 comp.realtime:3564 comp.os.386bsd.development:1074
Newsgroups: comp.periphs,comp.realtime,comp.os.386bsd.development
Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!convex!convex!darwin.sura.net!spool.mu.edu!olivea!hal.com!decwrl!netcomsv!netcom.com!jmonroy
From: jmonroy@netcom.com (Jesus Monroy Jr)
Subject: Why fix the RTC (Real Time Control) for 386bsd
Message-ID: <jmonroyCBuLr1.7L3@netcom.com>
Keywords: RTC report
Organization: NETCOM On-line Communication Services (408 241-9760 guest)
Date: Mon, 16 Aug 1993 10:42:37 GMT
Lines: 60
The text that follows is self-explainitory,
except that this is only the first 40 lines of the
report.
The complete report is available at
etext.archive.umich.edu
In /pub/Zines/QIC-News look for rtc.report
=====================================================================
Reasons to fix the RTC.
At the request of Bill Jolitz and other, I have
written this report. I should note that this is NOT a request
for a Real-Time OS (Operating System). Also, I was a bit
unaware of the lack of work done in this area, RTC (Real-Time
Control). My assumption was that other had this knowledge, as
second-hand, just as John Sokol and I. Our knowledge came
from work on our commercial products, the printer-port sound
dongle and the printer-port GPIB [communicator]. So in a
sense I am apologizing for any misunderstandings which may
have occurred.
-----------o0o--------------
This report is 1280+ lines long.
Footnoted reference are at the bottom of the report.
-----------o0o--------------
Bill (Jolitz's) main comments to me were "if you want
the RTC (clock) fixed you have to have some good reasons and
you have to show me why.". That is the main purpose of this
report. The next paragraph is an outline on the report.
This is a breakdown of the sections below. The
sections are broken up by a set of triple dashed lines with
the section number in the center of the line. The first
section shows that time is an issue and planning is better
than post (or run-time) analysis. The second section shows
that interrupts and the error recovery process are the main
contentions for efficient context switching and kernel
operations. So, not having a failure will greatly increase
system efficiency (This should be a Given.). Section Three
answer the popular myths about RTC. In section four are
comments from Mr. David Brown of UCSD about his
experiences with the QIC-40/80 implementation. Lastly in
Section Five are my comments and recommendations.
___________________________________________________________________________
Jesus Monroy Jr jmonroy@netcom.com
/386BSD/device-drivers /fd /qic /clock /documentation
___________________________________________________________________________