[9fans] Reinstalled system won't boot
Adrian Tritschler
on a KFS system. Since I needed more space for the Windows partition,
and wanted to convert the Plan9 system to fossil, I deleted the plan9
partition, then booted off a recent (this month) install CD and created
a 2.5G partition at the end of the disk.
Installation all went smoothly, partitions of 9fat, nvram, fossil and
swap. I selected "plan9" as the boot method and it replaced my MBR as
expected.
Attempting to reboot gets to:
MBR...PBS...Plan 9 from Bell Labs
ELCR: 0800
apm ax=f000 cx=f000 dx=40 di=ffff ebx=43b3 esi=-1
dev A0 port 1F0 config 045A capabilities 0F00 mwdma 0007 udma 041F
and then hangs. No combination of keystrokes I've tried seems to do
anything.
Booting from a floppy disk worked _once_, then I managed to do something
wrong while trying various "try this and try thats" from old 9fans and
screw up booting completely.
I've booted off the install CD, deleted the plan9 partition, recreated
it, reinstalled from the CD, and reselected the "plan9" boot.
Booting from the HD again gets me the above errors.
Booting from the floppy now gives the following:
1. sdC0!9fat!9pcf
2. sdC0!9fat!9pcf
3. none of the above
bootfile:
When I select either of "1" or "2"
root is from (tcp, il, local)[local!#S/sdC0/fossil]:
user[none]: glenda
time...
fossil(#S/sdC0/fossil)...version...boot: couldn't bind $rootdir=/root/3e
to root: '/root/3e' file does not exit
boot: second bind /: '/root/3e' file does not exist
panic: boot process died: unknown
dumpstack disabled
cpu0: exiting
Where to now?
Adrian
---------------------------------------------------------------
Adrian Tritschler mailto:aj...@ajft.org
Latitude 38°S, Longitude 145°E, Altitude 50m, Shoe size 44
---------------------------------------------------------------
Adrian Tritschler
dd is my friend.
Reinstalling from the CD seemed to see the existing plan9 partitions and
not rewrite 9fat or anything else.
Zeroing the entire partition and reinstalling, whilst crude, was quite
effective.
> Installation all went smoothly, partitions of 9fat, nvram, fossil and
> swap. I selected "plan9" as the boot method and it replaced my MBR as
> expected.
Now I'm back to this.
I can boot from a floppy.
I've installed the MBR that seems to have worked for me in the past, I
can select to boot either from my WinXP or Plan9 partition. Windows
still works, Plan 9 goes through the following and then hangs.
> MBR...PBS...Plan 9 from Bell Labs
> ELCR: 0800
> apm ax=f000 cx=f000 dx=40 di=ffff ebx=43b3 esi=-1
> dev A0 port 1F0 config 045A capabilities 0F00 mwdma 0007 udma 041F
Any suggestions on making the plan9 system boot from the harddisk?
Boyd 'rounin' Roberts
I remember the time we fixed a broken 32V file-system's /etc with a screwdriver, sticky tape, adb, dd and a magtape.
--
MGRS 31U DQ 52572 12604
Tom Ivar Helbekkmo
> I remember the time we fixed a broken 32V file-system's /etc with a
> screwdriver, sticky tape, adb, dd and a magtape.
That sounds outrageous enough to be true... Would you share the story?
-tih
--
Don't ascribe to stupidity what can be adequately explained by ignorance.
j...@plan9.bell-labs.com
You had a screwdriver?
Boyd 'rounin' Roberts
well, i had just coded autoreboot into 32V and my office mate [John Mackin]
booted the new kernel with the old init. that crashed and burned. so we
resorted to the stand alone stuff in /stand. trouble was that stand alone
fsdb was broken in such a way that you couldn't just nominate an inode
to look at. we were going to binary chop till we found /etc -- no chance.
so we found a screwdriver [jmk: the tool, not the drink, but i sure could
have slammed down a few at that point] and tape to tape it down on the
LA-120's return key and waited ... it took a long time [hours]. we'd
found that return would scroll you on to the next inode. iirc the tape
became unstuck serveral times and we had to restart.
once we had found /etc's inode we found the inode numbers for the
inits. we then stand alone dd'd the first block of /etc onto a 9 track,
mounted on a TU-16.
having another 11/780 with a 9 track we dd'd off the block, poked
around with adb -w and patched /etc/init's inode # to be the new init.
dd'd that block back onto the tape and dd'd it off onto the broken
vax.
booted it up and everything was fine.
john had assumed that the old init would work with the new kernel.
i was going to swap inits & kernels and boot sometime during the
early morning, but he beat me to it.
we had plans with removable RM-03 packs and other plans
that were just not going to work. it was sporty. some 128 of the
2000 students were not going to be happy, not to mention the
lecturing staff.
john mackin: http://www.std.org/~shand/jjm
Adrian Tritschler
gum, does anyone have any suggestions on what to do next when a fossil
installation on an IBM T21 can be booted from floppy disk, but hangs
when attempting to boot from the harddisk.
Windows XP (16G partition 1) still works, Plan 9(3G partition 2) goes
through the following and then hangs.
> MBR...PBS...Plan 9 from Bell Labs
> ELCR: 0800
> apm ax=f000 cx=f000 dx=40 di=ffff ebx=43b3 esi=-1
> dev A0 port 1F0 config 045A capabilities 0F00 mwdma 0007 udma 041F
locks up here and will only respond to the power switch.
Any suggestions on making the plan9 system boot from the harddisk?
Adrian
Richard Miller
Does the disk have a DOS or 9fat partition holding a plan9.ini file?
-- Richard
Adrian Tritschler
>>Any suggestions on making the plan9 system boot from the harddisk?
> Does the disk have a DOS or 9fat partition holding a plan9.ini file?
It has the partitions that the install CD offered to create, in the size
and order that it created them. There's a 16G NTFS and a 3G plan9
partition, with the plan9 partition divided up to include 9fat.
term% disk/fdisk -p /dev/sdC0/data
part ntfs 63 33264000
part plan9 33264000 39070080
term% disk/prep -p /dev/sdC0/plan9
part 9fat 33264000 33468800
part nvram 33468800 33468801
part fossil 33468801 38021504
part swap 38021504 39070080
Booting off the floppy it states that it is using the plan9.ini from the
9fat subpartition, and then the fossil system.
>using sdC0!9fat!plan9.ini
>found 9PCF . attr 0x0 start 0x4f9 len 2383465
:
>root is from (tcp, il, local)[local!#S/sdC0/fossil]:
> -- Richard
Richard Miller
>
> It has the partitions that the install CD offered to create, in the size
> and order that it created them.
It looks to me as if the 9fat partition is not being found by the
9load on your hard drive. Is it identical to the 9load on the
floppy which boots successfully? Can you build a 9load with
debugging output turned on (e.g. trace in part.c, parttrace
in devsd.c and chatty in dosboot.c), put that on 9fat and try
to boot with it?
-- Richard
Adrian Tritschler
Thanks for that.
I copied the 9load from the very old floppy (15-Feb-2002) to /n/9fat and
now I can boot the system from the hard disk. Maybe "in the future"
I'll dig around and try to find why the recent 9load won't load on this
system.
For anyone interested:
Booting is now via WinXP's boot manager, the NTFS partition is marked as
active, the C:\BOOT.INI contains a line to include "C:\Bootsect.p9" in
the start menu.