problems booting 386bsd Rel 0.0
Martin Grossman
"changing root device to fd0a" (some one mentioned they got fd0b?)
Whats up?
386DX-33 (18MHZ<->33MHZ)
8MB RAM
IDE -> ST1144A (disconnected so wont write over disk)
A: 3.5 1.4MB
B: 5.25 1.2MB
colorado 250MB tape
brand spanking new 387-33 (cyrix)
AMI BIOS of 4-9-90
64K cache
Paradise 1024K SVGA (WD)
2 serial ports
1 lineprinter port
1 ethernet card
SIS chipset
John Richards Rowland
the OPTI chipset? I have seen many reports of failure,
but there are also many OPTI boards. Also, anybody have
any ideas about an interpretation of the "Operator Abort-
386bsd not loaded, trying xxx, file xxx not found,trying yyy..."
error message? Which part of the system (i.e. not me) might
be referred to as the Operator?
--
-----------------------------------------------------------------------
primary: jrow...@cs.utexas.edu (UT CS Department)
secondary: jrow...@csdfx8a.arlut.utexas.edu (Applied Research Laboratory)
-----------------------------------------------------------------------
Roger Scott
>I also got as far as the line
>"changing root device to fd0a" (some one mentioned they got fd0b?)
I'm also in the same situation. I thought it may have been the serial
port cards, so I removed them. I then swapped the hard drive controller,
and finally the SVGA card. I still ended up with a panic. Surprisingly
both hard drives were still readable and for this I am thankfull I didn't
need to do a restore. Does anybody know when a version which may work will
be out? BTW this machine WILL run OS/2, Xenix and SCO Unix so why am I
having so many problems.
Setup:
386/33 4Mb Clone
64Kb cache - tried disabled and enabled
AMI BIOS
OPTI chipset
WD RLL & IBM MFM controller (I only tried these with a 30 & 20Mb drive)
ET3000 SVGA card & CGA card
Shadowing was disabled.
As you can see after many combinations it looks like my motherboard is the
problem:-( I don't suppose I should have given the software the 66Mb drive?
Roger.
*****************************************************************
* Roger Scott Honours Computer Science *
* University Of Tasmania, AUSTRALIA *
* Internet/AARnet u90...@bruny.cc.utas.edu.au (Try here 1st) *
* or ap...@cleveland.freenet.edu *
* or ro...@alchemy.tn.cornell.edu *
* or Roger...@p15.f206.n670.z3.fidonet.org *
*****************************************************************
Michael Bryan
>gros...@bbn.com (Martin Grossman) writes:
>
>>I also got as far as the line
>>"changing root device to fd0a" (some one mentioned they got fd0b?)
>
>I'm also in the same situation.
>be out?
According to Bill Jolitz, this problem is fixed in 0.1, due out sometime
next week, hopefully. The other common problem of "Operator abort"
messages is supposed to be fixed in the latest 0.0 (that's the *third*
release of 0.0, I think, which I haven't yet gotten myself.)
--
Michael Bryan mic...@resonex.com
This offer law where prohibited by void.
Mike Murphy
>Can anybody verify a boot by 386bsd on a board with
Yes. Works with a 386sx/387sx 16MHz, SVGA, IDE, Connor 3044, WD8003E, 1.2MB FD.
Didn't work with a Hercules clone, all other things the same. There was panic
after typing a character at the <color> string in the center of the screen after boot.
--
Mike Murphy m...@Optigfx.COM ucsd!optigfx!mrm +1 619 292 6060 x 265
Optigraphics Corporation 9339 Carroll Park Drive San Diego, CA 92121
The opinion(s) expressed above are mine and not those of my employer.
Michael Bryan
>These version numbers are quite hard to keep track on.
>We have had three versions of 0.0. The last one is also called 0.1.
>Now you say that next weeks version will also be called 0.1
To my knowledge, there have been no releases of "0.1". Anyone refering to
the 3/17/92 release as "0.1" is incorrect. There have been three releases
of "0.0", though. The first wasn't really intended for global release, and
was quickly followed by a "fixup" release, called "0.0new" by some. Then
another fixup version of 0.0 was released. According to Bill, this couldn't
be called 0.1, as there was something called 0.1 and 0.2 already in Beta
test at a few sites. So it stayed 0.0, and has been identified as "0.0newer"
and the "3/17" release of 0.0, that I've seen. (Possibly others, as well?)
I agree that it's confusing. I can understand the need for not moving up
to 0.1, if there is already something by that name in a testing stage.
I would like to see a consistent, easy-to-use identifying method for
releases that have the same basic version number, however. Two ideas that
come to mind are letter suffixes, such as "0.0", "0.0a", "0.0b", etc., or
dates: "0.0 3/12/92", "0.0 3/14/92", and "0.0 3/17/92". (I actually tend
to prefer the former, due to it's terseness...) It would definitely
help thwart confusion...
Bernard Steiner
> gros...@bbn.com (Martin Grossman) writes:
> >I also got as far as the line
> >"changing root device to fd0a" (some one mentioned they got fd0b?)
> I'm also in the same situation.
> 386/33 4Mb Clone
> 64Kb cache - tried disabled and enabled
> AMI BIOS
This sounds very familiar to me...
However, I've got an ET4000 VGA card and a WD1009 ESDI controller.
The system would not get going off a 5.25'' floppy. I just got odd register
dumps when pressing ALTSysRq F[678] - no clue what they mean.
The real surprise came with 3.5'' boot floppy. With no further interaction,
I got to about the same stage as with 5.25'' BUT got a panic on keypress.
Somehow I recalled that pressing 'd' during boot would get you into debug
mode. I did that, and - voila -
Operator abort blah blah
File not found blah blah
Unfortunately, the system then just behaved the same way as above.
BUT:
Much to my surprise, 386bsd actually booted when exactly the following
conditions were met:
shadow RAM disabled
proper HD drive type configured and HD connected
using 3.5'' boot floppy
pressing 'd' during boot and keeping it pressed till after the *second*
Operator abort... message
Anybody got a clue *why* ?
BTW is there some kind soul with 386bsd+sources up and running to provide
The Net with a floatless newfs ? Okay, so that's not a clean solution,
but nevertheless...
Cheers,
Bernard
--
Bernard Steiner, FB Informatik/IRB, Uni Dortmund, vox +49 231 755 2444
Postfach 500500, D-W-4600 Dortmund 50, Germany fax +49 231 755 2386
b...@Germany.EU.net ...!uunet!unido!bs
*III And they gave it Instructions, but knew it not. } From The Book of Nome,
*IV It is, they said, a Box with a Funny Voice. } Mezzanine v.III-IV
Richard Tobin
>To my knowledge, there have been no releases of "0.1". Anyone refering to
>the 3/17/92 release as "0.1" is incorrect.
The third version contains this string in the kernel:
386BSD Release 0.1 by William and Lynne Jolitz
So it's not surprising many people are calling it 0.1.
-- Richard
--
Richard Tobin,
AI Applications Institute, R.T...@ed.ac.uk
Edinburgh University.
Per Lindqvist
: In article <1992Mar21....@compuram.bbt.se> p...@compuram.bbt.se (Per Lindqvist) writes:
: To my knowledge, there have been no releases of "0.1". Anyone refering to
: the 3/17/92 release as "0.1" is incorrect.
documentation said it was 0.0. I never run the boot binaries, so I
cannot say.
: I agree that it's confusing. I can understand the need for not moving up
: to 0.1, if there is already something by that name in a testing stage.
: I would like to see a consistent, easy-to-use identifying method for
: releases that have the same basic version number, however.
The easiest is to call them 0.0.0, 0.0.1 and 0.0.2, and so on.
So we now have 4.3BSD version 0.0.2 :-)
I don't know who invented these ridiculus sub-version-number schemes.
UNIX used to be quite straigthforward in the beginning, version 6 and
version 7, etc. Somewhere, someone, started version III, and then
version V, after which came all sub-versions of V, and then
sub-sub-versions. It appears that the BSD world finally cought on to
this wonderful idea :-(
Now, what are BSDI calling their versions?
I seriously doubt that sources/binaries for 386BSD and BSD/386 will
stay compatible for very long. :-(
--
Per Lindqvist
Internet: p...@compuram.bbt.se Fidonet: Per Lindqvist @ 2:201/332
Chris Torek
BSD port ... but not yet. Today being Sunday, however, I felt like
taking a holiday from serious topics. :-) )
In article <1992Mar25....@compuram.bbt.se> p...@compuram.bbt.se
(Per Lindqvist) writes:
>I don't know who invented these ridiculus sub-version-number schemes.
>UNIX used to be quite straigthforward in the beginning, version 6 and
>version 7, etc. Somewhere, someone, started version III, and then
>version V, after which came all sub-versions of V, and then
>sub-sub-versions. It appears that the BSD world finally cought on to
>this wonderful idea :-(
(Actually, those are System III, etc., rather than Version III, etc.)
[DISCLAIMER: the following may be entirely apocryphal. Believe it or
not at your own risk.]
Originally, the Computer Science Research Group (CSRG) at Berkeley
numbered their releases more simply. The first BSD VAX system was
3BSD, and the next one was 4BSD. CSRG then later renamed this `4.0BSD'
so as to release 4.1BSD, a much-fixed but otherwise very similar system
to 4.0BSD. 4.1, then, was sort of a `patch'.
4.2BSD was originally supposed to be called `5BSD'. According to my
sources---whose veracity is unquestioned, since I never questioned
them :-) ---the name was left as `4.2', rather than being changed to
`5.0', due to our National Source of Delight: the Lawyers. (It is
strange that, given how everyone professes to hate lawyers, the U.S.
has a higher per capita lawyer count than any other country. Perhaps
it has something to do with our lawmaking body, the U.S. Senate, being
made up of politicians whose career backgrounds are, for the most part,
in law.... But never mind that.)
Apparently---so the story goes---if 4.2BSD were called `5BSD', the
lawyers would demand a completely new release contract. If, however,
4.2BSD were called `4.2BSD', nearly the same contract as used for
4.1BSD could be reused. The same held for 4.3BSD, 4.3-tahoe, and
4.3-reno: by using a `subversion' number, the lawyers would be
satisfied that each was merely a new version of an existing system.
So, the next time you cannot figure out which version of Unix you have,
blame the lawyers. :-)
--
In-Real-Life: Chris Torek, Lawrence Berkeley Lab CSE/EE (+1 510 486 5427)
Berkeley, CA Domain: to...@ee.lbl.gov