In article <1992Aug10.013816.27835@afterlife.ncsc.mil> dpkemp@afterlife.ncsc.mil (David P. Kemp) writes:
>In article <135@svb.guug.de> I wrote about "installboot"
>>
>># pwd
>>/usr/kvm/mdec
>># ./installboot -vl /boot bootsd /dev/rsd0a
>>Primary boot: bootsd
>>Secondary boot: /boot
>>Boot device: /dev/rsd0a
>>bootfile /boot not on bootdevice /dev/rsd0a
>RTEM - Read The Error Message. It says /boot not on bootdevice /dev/rsd0a.
It turned out that I should have said
# ./installboot -vl /mnt/boot bootsd /dev/rsd0a
with a _mounted_ file system sd0a(!) and the full path of the boot file in
that special situation.
Ok, but: Do you agree, that on _every_ fs there is a root? The manual
is _very_ misleading in this point, not to speak of the un-"unixish"
way of writing onto a mounted device via the raw device, this is ugly!!!
Every programm could find a file named /boot on a unmounted partion, and
that's what I'd expect with a command, where TFM says:
DESCRIPTION
The boot(8S) program is loaded from disk by bootblock code
which resides in the bootblock area of a disk partition. In
order for the bootblock code to read the boot program (usu-
ally /boot) it is necessary for it to know the block numbers
occupied by the boot program.
[stuff deleted]
bootfile is the name of the boot program, usually /boot.
You see? _Usually_ /boot! If this isn't misleading... Even better:
EXAMPLE
To install the bootblocks onto the root partition on a Xylo-
gics disk:
example% cd /usr/kvm/mdec
example% installboot -vlt /boot bootxy /dev/rxy0a
Arrrgghh. This sounds like installboot also does the copying stuff, which
it could do, by the way, if it insists on a mounted partition.
Ok, it's always one first time, now I know, but I don't like it, still.
After a moment of thinking it gets clearer: Track zero is not touched by
sync, so the fact that the fs is mounted is considered harmless.
Thanks to all of you out there who responded by email:
jw@owf.ch (Juerg Wanner)
Martin Frost <me@Sunburn.Stanford.EDU>
Mike.Sullivan@EBay.Sun.COM (Mike Sullivan {AKA Simon BarSinister})
From: Edward J. Hartnett <ejh@khonshu.Colorado.EDU>
(didn't record the answers over the net, all answers were basically the
same, also thank you very much!).
And a special remark to David (dpkemp@afterlife.ncsc.mil): You tried
to help me, I see. But I hate being treated like a fool! Nevertheless,
thank you anyway.
Regards
-svb.
-- Sebastian v. Bomhard <svb@guug.de>, Brodersen & Bomhard GmbH >>>> UNIX/Datenbanken/Netzwerke: Planung/Loesungen/Schulungen <<<< St.-Anna-Str. 13, D-W-8000 Muenchen 22 (Lehel), +49-89-298934
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:46 CDT