Install of silo fails on ultra-2

Install of silo fails on ultra-2

[Kris Boulez]

I just 'inherited' an ultra-2, which managment seems to
have forgotten about, and I wanted to do my 'good deed of the 
day', so I tried to install the ultrapenguin release on it.
This machine has no graphical card so I connect to the serial
port via another linux machine and minicom.
When booting with 'boot net serial' all works fine except when
at the end the install program asks to install SILO, but fails with
"mount failed: Device or resource busy".
Apparently I can't switch tty's with minicom to get at a shell prompt
to see what's happening ?
I did the same install on an other ultra (with graphical card) without
any problem

P.S. bootimage is tftp-ed and install via nfs mount.

Kris
-- 
Kris Boulez				kris@belbone.be

Re: Install of silo fails on ultra-2

[David S. Miller]

   Date: 	Wed, 27 May 1998 11:07:38 +0200
   From: Kris Boulez <kris@belbone.be>

   P.S. bootimage is tftp-ed and install via nfs mount.

I've seen this problem.  You can still salvage the installation by
booting the tftpboot image with root=/dev/xxx where xxx is the root
partition device.  Then once it is up, login as root and run
"/sbin/silo" then you can reboot from the disk just fine.

Later,
David S. Miller
davem@dm.cobaltmicro.com

Re: Install of silo fails on ultra-2

[Kris Boulez]

Quoting David S. Miller (davem@dm.cobaltmicro.com):
>    Date: 	Wed, 27 May 1998 11:07:38 +0200
>    From: Kris Boulez <kris@belbone.be>
> 
>    P.S. bootimage is tftp-ed and install via nfs mount.
> 
> I've seen this problem.  You can still salvage the installation by
> booting the tftpboot image with root=/dev/xxx where xxx is the root
> partition device.  Then once it is up, login as root and run
> "/sbin/silo" then you can reboot from the disk just fine.
> 
Booting with "boot net linux root=/dev/sda1" hangs at 


  Partition check:                                                              
   sda: sda1 sda2 sda4                                                          
  RAMDISK: Compressed image found at block 0                                    
  VFS: Mounted root (ext2 filesystem).                                          


Kris,
-- 
Kris Boulez				kris@belbone.be

Re: Install of silo fails on ultra-2

[David S. Miller]

   Date: 	Wed, 27 May 1998 11:45:53 +0200
   From: Kris Boulez <kris@belbone.be>

   Quoting David S. Miller (davem@dm.cobaltmicro.com):
   > I've seen this problem.  You can still salvage the installation by
   > booting the tftpboot image with root=/dev/xxx where xxx is the root
   > partition device.  Then once it is up, login as root and run
   > "/sbin/silo" then you can reboot from the disk just fine.
   > 
   Booting with "boot net linux root=/dev/sda1" hangs at 

Sorry, my fault.  Usually I also have a vmlinux.aout image in my
netboot install area too, and _this_ is what I use to recover from
failed silo invocations during an install.

I've built such an image that should hopefully work just fine for this
purpose.  Get it from:

ftp.cobaltmicro.com:/pub/users/davem/images/vmlinux.aout

Good luck.

Later,
David S. Miller
davem@dm.cobaltmicro.com

Re: Install of silo fails on ultra-2

[Kris Boulez]

> Sorry, my fault.  Usually I also have a vmlinux.aout image in my
> netboot install area too, and _this_ is what I use to recover from
> failed silo invocations during an install.
> 
> I've built such an image that should hopefully work just fine for this
> purpose.  Get it from:
> 
Got it and installed. Works only partially though :-( I get the following 
error

 zs_open cua0, tty overwrite.
 zs_open cua0, tty overwrite.
 zs_open cua0, tty overwrite.
 zs_open cua0, tty overwrite.
 zs_open cua0, tty overwrite.
 zs_open cua0, tty overwrite.

(I have a console attached to cua0). When I detach the console and attach 
it later I get the following oops

 
Unable to handle kernel NULL pointer dereference<1>tsk->mm->context = 0000000000000000
tsk->mm->pgd = fffff80000002000
              \|/ ____ \|/
              "@'/ .. \`@"
              /_| \__/ |_\
                 \__U_/
swapper(0): Oops
TSTATE: 0000009980f09600 TPC: 00000000005714a8 TNPC: 00000000005714ac Y: 01200000
g0: 0000000000413850 g1: 00000000ffffec00 g2: 0000000000000000 g3: 0000000000434e3c
g4: fffff80000000000 g5: 0000000000000000 g6: 0000000000410000 g7: 0000000000000000
o0: 0000000000000001 o1: 0000000000000000 o2: 00000000000000ff o3: 000000000000000d
o4: 0000000000000000 o5: 00000000000010c6 sp: 0000000000413111 ret_pc: 0000000000571480
l0: fffff8000fd78000 l1: 00000000005d7158 l2: 00000000000010c6 l3: 00000000005b77b0
l4: 00000000005d7158 l5: 0000000000000001 l6: 00000000005bb140 l7: 0000000000000008
i0: 000000000000000d i1: 0000000000662038 i2: 0000000000413ad0 i3: 00000000005712a0
i4: 00000000005b8af8 i5: 0000000000000000 i6: 00000000004131b1 i7: 00000000004176dc
Instruction DUMP: 90022001  d25c2238  d0242240 <c02a4000> 92026001  d05c2230  d2742238  f02a0000  90022001 
Aiee, killing interrupt handler
Kernel panic: Attempted to kill the idle task!
In swapper task - not syncing
Press L1-A to return to the boot prom


I got the machine booted one time cleanly and have silo installed now

Kris,
-- 
Kris Boulez				kris@belbone.be