home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.sgi
- Path: sparky!uunet!wupost!sdd.hp.com!mips!odin!fido!zola!zuni!anchor!olson
- From: olson@anchor.esd.sgi.com (Dave Olson)
- Subject: Re: Problems booting IRIX 4.0.4
- Message-ID: <oo41ht0@zuni.esd.sgi.com>
- Keywords: boot
- Sender: news@zuni.esd.sgi.com (Net News)
- Organization: Silicon Graphics, Inc. Mountain View, CA
- References: <1992Aug18.164824.23210@thunder.mcrcim.mcgill.edu>
- Date: Tue, 18 Aug 92 20:13:22 GMT
- Lines: 56
-
- In <1992Aug18.164824.23210@thunder.mcrcim.mcgill.edu> vasco@pet.mni.mcgill.ca (Vasco KOLLOKIAN) writes:
- | In a nutshell, our DataStation Server refuses to boot from IRIX 4.0.4 kernel
- | when turned off. We boot from IRIX 3.3.2 kernel (saved on another disk), reset
- | the machine because it doesn't go far, and afterwards successfully boots from
- | the regular IRIX 4.0.4 kernel which fsck's the disks and everything thereafter
- | proceeds normally.
-
- 3.3.X kernels have a different idea of the console driver/major # than
- 4.0.X kernels, so you really can't boot one on the other. What
- happens when you boot the 4.0.4 kernel? What is the console set to,
- and do you have an ascii terminal connected (presumably since it is
- a server, it doesn't have a graphics board, and therefore the console
- should default to d, but you never know...). Perhaps from inittab
- there is no login enabled on ttyd1, and something is mucked up in
- one of the boot scripts, or the entries in /dev may be wrong.
- Can the machine be accessed over the net; do the 4 LED's on the
- bottom of the CPU board blink after the "freeze"?
-
- Also note that the 4D/20 and 25 was NOT officially supported by 4.0.4,
- and received very limited testing. Even if the TAC shipped it to you
- for that particular machine (which may have happened), it STILL isn't
- supported (see earlier postings of mine on the distinction between
- "supported", and "may or may not work").
-
- I'd try 4.0.5, but see my other comments.
-
- | This problem is intriguing because we did a fresh installation only after the
- | entire system disk that contained IRIX 3.3.2 was copied (bru) to a spare disk
- | and the system disk itself erased before a fresh installation of 4.0.1 and
- | upgrade to 4.0.4 was done.
-
- Uh oh, I don't like the way that was stated. Did you boot the 4.0.4
- miniroot to install 4.0.4? If you used the 4.0.1 miniroot, you have
- a broken installation. Boot again from a 4.0.4 miniroot, do
- 'set neweroverride' once in inst, and reinstall eoe1 and eoe2, if so.
-
- | path=dksc(0,1,8) dksc(0,1,0)
- | sash: boot -f dksc(0,1,7)unix <-- boot from 4.0.4 kernel on disk 1
- | 1040816+77680+346640 entry: 0x80002000
- |
- | (freezes, and has to be reset)
-
- What does that mean? What is actually shown on the console. If nothing
- is shown at all, the kernel almost certainly is bad.
-
- | sash: boot -f dksc(0,1,0)unix
- | 1040816+77680+346640 entry: 0x80002000
-
- Are partition 0 and 7 really 2 completely distinct partitions? Normally
- as shipped, they start at the same place, but are different lengths;
- only one has a valid filesystem; I doubt that has any bearing on
- the problem, but it is curious that you show both.
- --
- Let no one tell me that silence gives consent, | Dave Olson
- because whoever is silent dissents. | Silicon Graphics, Inc.
- Maria Isabel Barreno | olson@sgi.com
-