On Tue 11 Jan, Russell King - ARM Linux Admin wrote:
> I guess that the zip drive is making an illegal SCSI state transition
> (most SCSI devices firmware is pretty poor, and it's virtually
> impossible to write a driver which will talk to everything if you don't
> have everything to test against).
OK. I'll get more details when I find my notes.
> > So, ignoring that glitch we continue and I get to the package selection
> > section (stage 6 IIRC). First I get an error 'unable to read header at
> > 604440'.
>
> That means that the installer is unable to read one of the package headers
> fron the RedHat/RPMS directory. This message is generated by a bit of
> over-simplified code - it should tell you the filename if possible.
>
> It sounds like one of the .rpm files is corrupted.
OK. So probably not fatal - but hard to fix withouth downloaded the whole
lot again or waiting till the installer tells me which one :-)
> > message it goes back to the select packages to install window, and I am
> > stuck going round and round in a loop, never being able to install any
> > packages.
>
> Does it say anything on the logging ttys?
OK, I've dine it again to get this right, and this time (despite having
done it at least 8 times already) I got slightly further than before:
...
VT 2 VT 0 (main install screen)
new stage 5 Header error: 'unable to read header at 604440'.
Coponent Warnings (29 of 459 packages...)
new stage 6 'Components to Install' chooser dialog
verifying dependencies
Dep: bash needs filutils
Dep: kernelcfg needs python
Dep: kernelcfg needs pythonlibs
Dep: newt needs slang
Dep: zsh needs fileutils
'Unresolved dependencies' box (giving these 5 deps)
I enter 'no' here.
new stage 7
net shared path is not set
'Install Log' dialog
unmounting all filesystems
formatting filesystems
mounting all filesystems
mounting sdb2 on /mnt/ as ext2
rpm database openned
pre-installing *beforeskel*
Installing *beforeskel*
Installing base
'Install of fpem failed'
Installing other package
rpm database closed
unmounted filesystems
new stage 6
Now we are back to 'the components to install' windows again and can only
go round in circles.
I downloaded a new copy of the fpem-2.0-1.arm.rpm but got the same
complaint. Is there something else other than the frp, rpm itself that
might cause this error? On previous attempts it hasn't managed to install
beforskel or base before - don't know why it did this time...
Anyway, having got this far I think I should be able to get it to boot
and can work from there...
> > System 2:
> > ROS 3.71, 17Mb (16+1), ARM 700 +FPA 10
> > Drives: motherboard IDE: 4Gb WD caviar 34000
> >
> > On this system I cannot boot from a floppy with
> > !linux -bootkernel ADFS::0
>
> That's expected - the name after -bootkernel is the RISC OS filename of
> the kernel. ADFS::0 is not. It won't report an error, because RISC OS
> does allow you to open and read from "ADFS::0".
>
> This would explain why your following attempts also failed.
Except that the above command still works fine on system 1, and is the
way I have sucessfully installed ARMLinux on previous occasions.
If I put the kernel and root images (and !linux) in the root of ADFS::4
and use
!linux -bootkernel kernel -initrd root
that hangs in the same way on system 2, and works fine on system 1. There
is clearly some other problem too...
Wookey
--
Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel (00 44) 1223 811679
work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
Wookey
--
Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel (00 44) 1223 811679
work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
++ Please use [EMAIL PROTECTED] for ++
++ kernel-related discussions. ++
