Return to BSD News archive
Newsgroups: comp.os.386bsd.misc
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!library.ucla.edu!news.mic.ucla.edu!unixg.ubc.ca!acs.ucalgary.ca!cpsc.ucalgary.ca!xenlink!fsa.ca!deraadt
From: deraadt@fsa.ca (Theo de Raadt)
Subject: Re: Will this work (two IDE drives, DOS and *BSD)?
In-Reply-To: adam@veda.is's message of 24 Aug 93 21: 46:02 GMT
Message-ID: <DERAADT.93Aug24170735@newt.fsa.ca>
Sender: news@fsa.ca
Nntp-Posting-Host: newt.fsa.ca
Organization: little lizard city
References: <CC73xF.8n0@rex.uokhsc.edu> <1993Aug23.164125.1497@fcom.cc.utah.edu>
<MYCROFT.93Aug23224109@trinity.gnu.ai.mit.edu>
<1993Aug24.200723.26095@fcom.cc.utah.edu> <CCA9t4.us@veda.is>
Date: Wed, 25 Aug 1993 00:07:35 GMT
Lines: 37
adam@veda.is (Adam David) writes:
> terry@cs.weber.edu (A Wizard of Earth C) writes:
> >
> >As to the load from another, drive, yes, Julian's boot blocks support you
> >typing in the drive and partition (if you already know it and enjoy typing
> >in that syntax) -- but you'd have to be the best typist in the world to get
> >it typed in on my machine in the timeout period before autoboot. My machine
> >gives almost 0 delay (50MHz EISA, 0 wait state, 9ms drive on 1742 SCSI).
>
> Isn't it enough to type _anything_ in the delay period and then you have as
> much time as you like for correcting the line before pressing the return key?
> Surely there is enough time to hit one of the keys (for instance, near the
> middle of the keyboard) before it auto-defaults to the first disk.
My 66MHz EISA machine exhibits the same behaviour. I only get about 500msec
to hit my first key (after which, yes, I can type at any speed I want). The
reason for this problem is because the boot code spinwaits, somewhat like this:
for(i=0; i<50000; i++) {
gateA20();
check key();
gateA20();
}
It's not that our machines are especially fast. It is because our
machines come with extremely fast hardware implimentations of the
FAST-A20 gate hack (one of the most disgusting things in the PC
architecture, which all machines from the AT and up have.)
There are fixes from <hpeyerl@fsa.ca> which fix this problem
(basically, instead of spinwaiting, his code watches the RTC clock
"tick") but they have not been merged into the distribution yet. The
boot blocks are getting pretty big, and it is becoming difficult to
add things to them..
--
This space not left unintentionally unblank. deraadt@fsa.ca