Date: 05 Aug 1998 20:12:44 +0200 From: Benedikt Stockebrand <benedikt@devnull.ruhr.de> To: Steve Roskowski <rosko@mpath.com> Cc: Benedikt Stockebrand <benedikt@devnull.ruhr.de>, freebsd-stable@FreeBSD.ORG Subject: Re: install friendly? Message-ID: <87zpdjti03.fsf@devnull.ruhr.de> In-Reply-To: Steve Roskowski's message of "Tue, 04 Aug 1998 17:23:10 -0700" References: <Steve Roskowski's message of "Mon, 03 Aug 1998 11:54:54 -0700"> <3.0.5.32.19980803115454.0155c100@mail.mpath.com> <3.0.5.32.19980804172310.01567cd0@mail.mpath.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Steve Roskowski <rosko@mpath.com> writes:
> the docs with the motherboard specifically suggest for configs with IDE HD,
> CDROM and OTHER that the OTHER be second device on the second bus.
Looks like those people writing those docs should be shot on sight ---
or rather, forced to read all their docs.
> I will
> move it & see. Not being an IDE wiz, it was unclear whether the term
> master and slave refered to ownership of the bus or to initiate/response
> relationship (as in bus master/bus slave on PCI...), so I went with the
> recommendation.
I'm not particularly into IDE things either, but I still remember that
the master device only deals with the controller card.
IIRC the historical reason is that MFM and RLL controllers had
separate control connectors to each drive and a shared data connector
to both. The IDE people moved most of the controller card logic into
the disk itself, leaving only the most rudimentary stuff on the
controller card. Now having two disks on an IDE controller gets you
two of these disk-side controller parts and you only need one. The
slave simply turns its controller off. But if you don't have a master
device you don't have a full controller, just that stub on the card.
So much about the amazingly innovative technology of the IDE bus.
Someone pass me an ST01 please...
> The frustration with the whole install stuff was I thought I followed
> directions, and did exactly what the novice install procedure suggested.
> Working with the 2.2.7 CD from Walnut Creek, built a floppy and went for it.
You'll have at least as much trouble if you try to run NT and any
other operating system on the same box but from different drives.
> step 1) partitioned the scsi drive to MBR (no boot loader) and the standard
^^^^^^^^^^^^^^^^^^^^^^^
I don't understand what you mean here.
> 3 FreeBSD partitions via the (A) command (use all the disk).
> step 2) went to the IDE drive and tried to put the bootloader on it. This
> was pretty confusing, since the only way to touch it was to say I wanted to
> partition it then quit, but when it asked I said yes, put a bootloader on it.
Hmm. Guess I'll stay with the "dangerously dedicated" disklabels on my
boxes :-)
No, you don't really have to understand that. But there's a bit of a
problem with the way disks, partitions and slices are handled with all
the PeeCee BSDs. If you really want to find out you might search the
docs about this subject.
> step 3) did the rest of the install (all from 2.2.7 CDROM)
> step 4) reboot the system
> step 5) pops up to the boot loader, I say BSD
> step 6) BSD starts initial stuff, I get a boot prompt and let it timeout
> step 7) BSD panic stops, can't find root, reboot
Right, but that's not the BSD panicking but the boot loader. It tries
to boot from the first _BIOS_ disk using the usual BIOS I/O routines.
But since that first disk is your Win95 disk it doesn't find a BSD
partition to boot from and panics.
> step 8) system reboots, never gets beyond the bios, no valid HDD
Right. The first disk hasn't got a Windows MBR anymore and you don't
tell it to boot from the second disk.
> step 9) put the BSD boot floppy back in & reboot (stupidly I did not have a
> Win95 boot disk lying around)
If people were born smart we couldn't charge them such ridiculous
money for fixing the mess they made :-)
> step 10) look @ disk configuration (via config/fdisk option) and IDE drive
> looks like it has been converted into 3 partition, MBR, FreeBSD (big),
> unknown (small) at the top (this is from memory, so the names may be off).
Hmm. Can't figure out what's going on there.
> step 11) try to boot directly to SCSI (by changing bios option), again panics
That's probably because you haven't got a proper MBR on that disk.
> Ok, but hardly intuitive, and the boot
> prompt goes by really fast on this machine (like 10 seconds at most on 350
> Mhz Pentium II), NOT enough time to read & grok the long list of
> instructions about how to make it boot to some other drive.
Press space when the boot prompt shows up. That'll stop the timeout
and you'll have all the time it takes to read whatever is coming up.
> naively, I expected the novice install to correctly configure the boot
> model to match what I had installed through the install tools.
You'll always get into some kind of trouble with dual-boot setups.
Usually there's a way to figure things out, but when I wrote my master
thesis I used a switch to turn the power supply of some SCSI disks off
(that was the one with my thesis on, so I couldn't mess it up when I
was tweaking my system). Later on this switch turned out to be _very_
useful for running multiple OSes on a single box without unnecessary
hassle. Unfortunately it requires SCSI disks; at least some IDE disks
don't shut down when the power connector is turned off but get their
power from the data connector instead.
So long,
Ben
--
Ben(edikt)? Stockebrand Un*x SA
My name and email address are not to be added to any list used for advertising
purposes. Any sender of unsolicited advertisement e-mail to this address im-
plicitly agrees to pay a DM 500 fee to the recipient for proofreading services.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?87zpdjti03.fsf>