Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 2 Mar 1995 21:24:34 -0500 (EST)
From:      Wankle Rotary Engine <wpaul@skynet.ctr.columbia.edu>
To:        robsch@robkaos.ruhr.de (Robert Schien)
Cc:        bugs@FreeBSD.org
Subject:   Re: 021095-SNAP
Message-ID:  <199503030224.VAA02233@skynet.ctr.columbia.edu>
In-Reply-To: <m0rkDGA-00098BC@robkaos.ruhr.de> from "Robert Schien" at Mar 2, 95 10:57:49 am

index | next in thread | previous in thread | raw e-mail

They say this Robert Schien person was kidding when he wrote:
> 
> > 
> > On Wed, 1 Mar 1995, Michael Quigley wrote:
> > 
> > > I've not been able to boot the latest snapshot version on at least 3 
> > > machines at my disposal.  They are all of varying configurations and 
> > > processors.  Basically what happens, is that they lock up when trying to 
> > > boot off the floppy.  I've made my boot disk the same way I did with 
> > > 2.0-RELEASE:
> > > 
> > > dd if=<filename> of=/dev/fd0 bs=18k
> > > 
> > > With no luck.  Any ideas?
> > 
> > I had this same thing happen to me.  What always get me by it (this may 
> > be totally different from what's happening with you) is to repeatedly tap 
> > the num lock key while it's booting.
> > 
> > ie. hit turn on the computer, right before it hits the floppy start 
> > tapping num lock until you see the FreeBSD boot prompt.
> > 
> > Don't ask me why this works, I have no clue, but it DOES work.  My 
> > computer would not boot (it would lock) any I did this and it fixed it.  
> > Then I tried it again to make sure it wasn't a fluke.
> > 
> I have made a similiar experience.
> I have to reset my machine a few times before FreeBSD will be booted.
> Never had this problem with 2.0-RELEASE.
> Why was the boot code changed?
> 
> Robert
> 

The bootstrap was changed to allow the system to automatically boot
with the serial port as a console if the keyboard isn't plugged in.
This is how Sun machines work: the boot prom looks for a keyboard
device and if it can't find one, it uses serial port A as the console.
This is useful for dedicated file or compute servers that don't need
any graphics hardware. Originally, the only way to get a serial
console in FreeBSD was to compile a special kernel with
'options COMCONSOLE', and doing this would always force the system
into 'serial console mode' whether you liked it or not. With 
FreeBSD-current, you can still use 'options COMCONSOLE', but you
can also get a serial console by booting with the -h flag or just by 
booting with your keyboard unplugged. This allows you to stick a
dumb terminal with a null modem on COM1 and use your system without
a keyboard, monitor, or even a graphics card. 

The problem is that unlike Sun workstations and servers, keyboards and
keyboard controllers tend to vary from one PC to another. This makes
it hard to come up with a foolproof keyboard probe that will work on
all machines *and* that will fit comfortably inside the tiny space
available in the bootstrap code. I originally started with essentially
the same keyboard probe that syscons uses, but it turned out a little
big so I stripped it down to the bare essentials. This worked on all
the hardware I tried it on, including a Gateway P90 PCI system with a
PS/2 mouse. I thought I had lucked out. Now I'm not so sure.

What the probe does is try to reset the keyboard. If the reset
succeeds, the keyboard is assumed to be present. If the reset
fails, the bootstrap uses COM1 as the console I/O device. It also
tells the kernel to use COM as the console instead of ttyv0.
Either way the system is supposed to boot; you just won't be able
to see anything if the probe mistakenly decides that your keyboard
isn't there.

If the system actually hangs (there's no more floppy activity
or anaything) then it's possible that the keyboard probe is wedging
the machine. What I have done to try to fix this for now is to
put back the rest of the keyboard probe code from syscons that I
originally took out. This still works on my hardware, but it
also fills up almost all of the room left in the boot blocks.
If this next attempt still screws up with some systems, then
either it'll have to be taken out or someone will have to come
up with a less invasive and more reliable probe.

I'm looking forward to the day when we finally have a three stage
bootstrap with a /boot program that gives us to room to put in
more elaborate features and bulletproofing.

-Bill

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~T~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-Bill Paul            (212) 854-6020 | System Manager
Work:         wpaul@ctr.columbia.edu | Center for Telecommunications Research
Home:  wpaul@skynet.ctr.columbia.edu | Columbia University, New York City
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The Møøse Illuminati: ignore it and be confused, or join it and be confusing!
~~~~~~~~ FreeBSD 2.1.0-Development #0: Tue Feb  7 01:49:07 EST 1995 ~~~~~~~~~


help

Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199503030224.VAA02233>