Subject: Re: Boot block oddness...
To: None <email@example.com,>
From: Matthias Drochner <firstname.lastname@example.org>
Date: 09/01/1998 10:40:58
Excerpts from netbsd: 31-Aug-98 Boot block oddness... Mason Loring
> /user/mdec/installboot -v /usr/mdec/biosboot.sym /dev/rwd0a
> failed... Or, rather, I got through it, but then had a system that wouldn't
> boot. <cheer>
There were already two reports about -current boot failures,
but none of them containes enough details to be useful for
tracking down the problem.
To start, did you get a printout (eg "No bootable partition")
when the boot failed?
> Being rash and impulsive, I tried running /usr/mdec/installboot (the -current
> version) from my /usr partition, and that failed, probably because it's built
> under UVM and -current, etc.
-current "installboot" with a 1.3.2 kernel? This might fail due to
system call changes... Would be better if you mentioned the
> 1) Does anyone know why the -current (built today!) installboot fails? Is
> it something I did, possibly?
There are no functional changes in "installboot" between 1.3.2
and -current. The only changes were for "lint" and the replacement
of the deprecated "mktemp()". Perhaps I've botched something
in this process...
Can you run both versions with the "-v" option and report about
> 2) Are there any inherent dangers in using a 1.3.2 installboot with a set
> of -current boot blocks, as I did?
No - as I said, the "installboot"s are supposed to function indentically.
> 3) Can someone tell me how exactly I make a bootable floppy out of my
> -current stuff?
If you start with a formatted floppy:
newfs /dev/rfd0a floppy3
/usr/mdec/installboot -f -v /usr/mdec/biosboot.sym /dev/rfd0a