LILO 01 01 01 01 01 01....

LILO 01 01 01 01 01 01....

Post by De Clark » Mon, 15 Mar 1999 04:00:00



Please send me mail if you have any ideas about this
sad, sad story.

I have a LILO boot problem...

System:  P133 Advanced/EV board, NCR 53C825 PCI SCSI ctrlr,
96 MB RAM, SCSI boot disk sda, SCSI cdrom.  This system has
been running *only* Linux for about 2.5 years, through a
few new RH releases.  It had 32 MB until just last week :-)

Recently I added 2 4GB IDE drives, which have been working
fine for about 3 weeks.

Then I added 2 32MB SIMMS to bring the memory complement
up to 96MB.  But Linux only saw 64 MB.  I left that
little problem for later...

Then I got brave and tried installing an Evergreen Spectra
333 CPU upgrade.  Evergreen expects this upgrade to work
on my Intel motherboard.

The upgrade required replacement of the old AMI BIOS with
MR BIOS.  Did that, and things seemed to go on working just
fine.  Then I made the classic error of doing 2 things
at once :-(

I asked an Intel hw hacker about the memory issue and on his
advice I modified lilo.conf to add append="mem=96m" and also
shut down to install the new CPU.  When I rebooted with the
new CPU I got as far as

        LI_

and that was it.

I thought the Evergreen CPU didn't work.  So...
put the old CPU back in and rebooted:

        LI_

and that was it :-(

I tried booting RH install diskette in the thought that I
could just reinstall the OS on the boot disk.  RH installer
can't find the cdrom device on the SCSI bus.

After a few iterations trying to get RH to see the CDROM,
the boot behaviour changed to

        [something, probably some portion of LILO, goes by
         too fast to read] 01 01 01 01 01 01 01 01 01 01 01
        01 01 01 01 01 01 01 01 01 01 01 01....

you get the idea.  The 01 text is just pouring onto the
screen at max scroll speed.

So... ummm... I don't have a rescue diskette... I just
upgraded to RH 5.1 recently and didn't get around to making
it yet [yes, this is very embarrassing].

Does anyone understand (a) what 01 01 01 01 means
                 (b) why I can't see my CDROM any more
                 (c) how I can ever get this crate booted again
I tried typing in "linux single root=/dev/sda1" at the
boot prompt on the RH install disk but that, obviously,
didn't work... it does mount the ext2 fs on the SCSI disk,
which gives me *hope* that the disk is still alive;  but
I can't get a shell at all.

Is there a simple hardware diag package on floppy disk that
I could run to find out at least how much of my system
is still alive?

The NCR SCSI card is alive enough to report SCSI ID strings
during initialization.  The rest of the system is alive
enough that MR BIOS shows correct settings for memory
size, CPU vendor and ID, etc.  There are really only 3
symptoms:

I can no longer boot from my SCSI boot disk
I can no longer even SEE my SCSI CDROM from RH installer
        so I can't reinstall

and one "operator error"

I don't have a rescue disk [stupid me!]

Is my LILO problem due to the append statement I added
or due to the death of my scsi controller, and how can
I tell?  And is there any way to get this thing booted
without its own special custom rescue disk?

de

--
.............................................................................
:De Clarke, Software Engineer                     UCO/Lick Observatory, UCSC:

:Web: www.ucolick.org | ones people hate, and the ones people don't use.--JO:

 
 
 

LILO 01 01 01 01 01 01....

Post by Justin B Willough » Mon, 15 Mar 1999 04:00:00


Use dejanews.com this question/problem has been asked before.

- Justin


> Please send me mail if you have any ideas about this
> sad, sad story.

> I have a LILO boot problem...

> System:  P133 Advanced/EV board, NCR 53C825 PCI SCSI ctrlr,
> 96 MB RAM, SCSI boot disk sda, SCSI cdrom.  This system has
> been running *only* Linux for about 2.5 years, through a
> few new RH releases.  It had 32 MB until just last week :-)

> Recently I added 2 4GB IDE drives, which have been working
> fine for about 3 weeks.

> Then I added 2 32MB SIMMS to bring the memory complement
> up to 96MB.  But Linux only saw 64 MB.  I left that
> little problem for later...

> Then I got brave and tried installing an Evergreen Spectra
> 333 CPU upgrade.  Evergreen expects this upgrade to work
> on my Intel motherboard.

> The upgrade required replacement of the old AMI BIOS with
> MR BIOS.  Did that, and things seemed to go on working just
> fine.  Then I made the classic error of doing 2 things
> at once :-(

> I asked an Intel hw hacker about the memory issue and on his
> advice I modified lilo.conf to add append="mem=96m" and also
> shut down to install the new CPU.  When I rebooted with the
> new CPU I got as far as

>    LI_

> and that was it.

> I thought the Evergreen CPU didn't work.  So...
> put the old CPU back in and rebooted:

>    LI_

> and that was it :-(

> I tried booting RH install diskette in the thought that I
> could just reinstall the OS on the boot disk.  RH installer
> can't find the cdrom device on the SCSI bus.

> After a few iterations trying to get RH to see the CDROM,
> the boot behaviour changed to

>    [something, probably some portion of LILO, goes by
>     too fast to read] 01 01 01 01 01 01 01 01 01 01 01
>    01 01 01 01 01 01 01 01 01 01 01 01....

> you get the idea.  The 01 text is just pouring onto the
> screen at max scroll speed.

> So... ummm... I don't have a rescue diskette... I just
> upgraded to RH 5.1 recently and didn't get around to making
> it yet [yes, this is very embarrassing].

> Does anyone understand (a) what 01 01 01 01 means
>             (b) why I can't see my CDROM any more
>             (c) how I can ever get this crate booted again
> I tried typing in "linux single root=/dev/sda1" at the
> boot prompt on the RH install disk but that, obviously,
> didn't work... it does mount the ext2 fs on the SCSI disk,
> which gives me *hope* that the disk is still alive;  but
> I can't get a shell at all.

> Is there a simple hardware diag package on floppy disk that
> I could run to find out at least how much of my system
> is still alive?

> The NCR SCSI card is alive enough to report SCSI ID strings
> during initialization.  The rest of the system is alive
> enough that MR BIOS shows correct settings for memory
> size, CPU vendor and ID, etc.  There are really only 3
> symptoms:

> I can no longer boot from my SCSI boot disk
> I can no longer even SEE my SCSI CDROM from RH installer
>    so I can't reinstall

> and one "operator error"

> I don't have a rescue disk [stupid me!]

> Is my LILO problem due to the append statement I added
> or due to the death of my scsi controller, and how can
> I tell?  And is there any way to get this thing booted
> without its own special custom rescue disk?

> de

> --
> .............................................................................
> :De Clarke, Software Engineer                     UCO/Lick Observatory, UCSC:

> :Web: www.ucolick.org | ones people hate, and the ones people don't use.--JO:

--
   _/     _/_/_/  _/    _/  _/    _/ _/   _/   RULES!!!!!!! * LINUX RULES *
  _/       _/    _/_/  _/  _/    _/   _/_/     Justin Willoughby
 _/       _/    _/  _/_/  _/    _/     _/      http://www.nmc.edu/~willouj/
_/_/_/ _/_/_/  _/    _/  _/_/_/_/    _/ _/     ------ Jesus Is Lord ------

 
 
 

LILO 01 01 01 01 01 01....

Post by De Clark » Mon, 15 Mar 1999 04:00:00



Quote:> Use dejanews.com this question/problem has been asked before.
> - Justin

Yes, I know.  I have read everything I could find on LILO
boot hangs in dejanews for the last 2 months :-)

My circumstances are not exactly like others I have seen
there.  I have also not seen any real explanation of the
01 01 01 behaviour.  What does it really mean?

I have tried sneaking in via RH install boot diskette using

        boot:  linux single root=/dev/sda1 initrd=

but the boot could not find the scsi disk,  in fact it found
0 targets on my scsi bus, despite the fact that the NCR
scsi adapter displayed correct scsi IDs for both targets
at power up.  Since the other target is my cdrom I'm really
hosed... can't even reinstall Red Hat from scratch.

Should I be suspecting that the scsi adapter has died now,
on top of everything else?  do they die that easily (just a
few power cycles)?  are there any low level hardware diags
for intel hardware other than WinDoze?

de

 
 
 

LILO 01 01 01 01 01 01....

Post by Patrik Magnuss » Mon, 15 Mar 1999 04:00:00





|>
|> > Use dejanews.com this question/problem has been asked before.
|>
|> > - Justin
|>
|> Yes, I know.  I have read everything I could find on LILO
|> boot hangs in dejanews for the last 2 months :-)
|>
|> My circumstances are not exactly like others I have seen
|> there.  I have also not seen any real explanation of the
|> 01 01 01 behaviour.  What does it really mean?

According to /usr/doc/lilo/Manual.txt:

Disk error codes
- - - - - - - -

If the BIOS signals an error when LILO is trying to load a boot image, the
respective error code is displayed. The following BIOS error codes are
known:

   0x00   "Internal error". This code is generated by the sector read
    routine of the LILO boot loader whenever an internal inconsistency is
    detected. This might be caused by corrupt files. Try re-building the
    map file. Another possible cause for this error are attempts to access
    cylinders beyond 1024 while using the LINEAR option. See section "BIOS
    restrictions" for more details and for how to solve the problem.
   0x01   "Illegal command". This shouldn't happen, but if it does, it may
    indicate an attempt to access a disk which is not supported by the
    BIOS. See also "Warning: BIOS drive 0x<number> may not be accessible"
    in section "Warnings".

|> I have tried sneaking in via RH install boot diskette using
|>
|>   boot:  linux single root=/dev/sda1 initrd=
|>
|> but the boot could not find the scsi disk,  in fact it found
|> 0 targets on my scsi bus, despite the fact that the NCR
|> scsi adapter displayed correct scsi IDs for both targets
|> at power up.  Since the other target is my cdrom I'm really
|> hosed... can't even reinstall Red Hat from scratch.
|>
|> Should I be suspecting that the scsi adapter has died now,
|> on top of everything else?  do they die that easily (just a
|> few power cycles)?  are there any low level hardware diags
|> for intel hardware other than WinDoze?
|>
|> de
|>

 
 
 

LILO 01 01 01 01 01 01....

Post by Jean-Serge Gagno » Mon, 15 Mar 1999 04:00:00


I had a similar problem with one SCSI and one IDE drive. If I ran LILO
with the IDE drive connected, LILO would complain that the /dev/sda was
not the first drive, but would still run (even if my BIOS said to use
SCSI as boot drive instead of IDE [this is windoze stuff]), except that
when I reboot, I get the LI error...

My solution is that I need to unplug the IDE drive (yes unplug, just
telling the BIOS that the 'C' drive is 'none' is not good enough, Linux
still finds 'hda') and then boot with a rescue disk and run:

mount /dev/sdaX /mnt ; lilo -r /mnt ; umount /mnt

After that, I can shutdown and reconnect the IDE drive and then
everybody's happy...

I have not been able to find references to this 'SCSI-IDE-LILO' problem
(haven't looked that hard)

I keep getting the error again once in a while cause I forget to
disconnect my IDE before running LILO, so I keep a rescue disk handy...

Hope this helps

P.S. Make sure you check your BIOS for SCSI/IDE startup options





> |>
> |> > Use dejanews.com this question/problem has been asked before.
> |>
> |> > - Justin
> |>
> |> Yes, I know.  I have read everything I could find on LILO
> |> boot hangs in dejanews for the last 2 months :-)
> |>
> |> My circumstances are not exactly like others I have seen
> |> there.  I have also not seen any real explanation of the
> |> 01 01 01 behaviour.  What does it really mean?

> According to /usr/doc/lilo/Manual.txt:

> Disk error codes
> - - - - - - - -

> If the BIOS signals an error when LILO is trying to load a boot image, the
> respective error code is displayed. The following BIOS error codes are
> known:

>    0x00   "Internal error". This code is generated by the sector read
>     routine of the LILO boot loader whenever an internal inconsistency is
>     detected. This might be caused by corrupt files. Try re-building the
>     map file. Another possible cause for this error are attempts to access
>     cylinders beyond 1024 while using the LINEAR option. See section "BIOS
>     restrictions" for more details and for how to solve the problem.
>    0x01   "Illegal command". This shouldn't happen, but if it does, it may
>     indicate an attempt to access a disk which is not supported by the
>     BIOS. See also "Warning: BIOS drive 0x<number> may not be accessible"
>     in section "Warnings".

> |> I have tried sneaking in via RH install boot diskette using
> |>
> |>      boot:  linux single root=/dev/sda1 initrd=
> |>
> |> but the boot could not find the scsi disk,  in fact it found
> |> 0 targets on my scsi bus, despite the fact that the NCR
> |> scsi adapter displayed correct scsi IDs for both targets
> |> at power up.  Since the other target is my cdrom I'm really
> |> hosed... can't even reinstall Red Hat from scratch.
> |>
> |> Should I be suspecting that the scsi adapter has died now,
> |> on top of everything else?  do they die that easily (just a
> |> few power cycles)?  are there any low level hardware diags
> |> for intel hardware other than WinDoze?
> |>
> |> de
> |>

 
 
 

LILO 01 01 01 01 01 01....

Post by Bob » Tue, 16 Mar 1999 04:00:00


It may be totally unrelated because my stuff is nothing like
yours but  the 01 01 01 problem seems strangely similar to when
I first booted up RH 5.2 a fortnight ago. What I got locked
like:
lo
lo
lo
.(that's what it looked like, anyway).

rebooting from the rescue disk started the machine up fine I
eventually traced the problem to the lilo section in linuxconf -
setting the linear read on did the trick.

I had scsi probs, too but I can't offer anything very relevent.
If it was an adaptec I'd suggest EZ-scsi).

Cheers
Bob

 
 
 

LILO 01 01 01 01 01 01....

Post by Alan Wes » Tue, 16 Mar 1999 04:00:00


I just spent a week trying to sort mine out, I was getting all these 01 01
01 just constantly appearing on the screen, I can't install anything on that
hard disk now tried deleting all partitions re-formatting it, and putting NT
back on it and it still comes up with the 01's. So basically now I have a
hard disk that'll never boot again. All because LILO tried to stick itself
in the MBR.

Oh and at the same time it managed to damage the hard drive I used to
install Linux from... Looks like I'll have to download it all again, great
eh?

 
 
 

LILO 01 01 01 01 01 01....

Post by De Clark » Sat, 20 Mar 1999 04:00:00


I've been doggedly pursuing my LILO problem.  The next logical
step was to check out the hardware.

I put the SCSI boot disk on another machine's scsi bus (known
to work) and took a look at it with fdisk:  odd results!

Disk /dev/sdc: 255 heads, 63 sectors, 263 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start      End   Blocks   Id  System

/dev/sdc1   *         1        9    66496+  83  Linux native
Partition 1 has different physical/logical endings:
     phys=(31, 65, 63) logical=(8, 71, 63)
Partition 1 does not end on cylinder boundary:
     phys=(31, 65, 63) should be (31, 254, 63)

/dev/sdc2             9       26   135135   82  Linux swap
Partition 2 has different physical/logical beginnings (non-Linux?):
     phys=(32, 0, 1) logical=(8, 72, 1)
Partition 2 has different physical/logical endings:
     phys=(96, 65, 63) logical=(25, 26, 63)
Partition 2 does not end on cylinder boundary:
     phys=(96, 65, 63) should be (96, 254, 63)

/dev/sdc3            26       91   525987   83  Linux native
Partition 3 has different physical/logical beginnings (non-Linux?):
     phys=(97, 0, 1) logical=(25, 27, 1)
Partition 3 has different physical/logical endings:
     phys=(349, 65, 63) logical=(90, 149, 63)
Partition 3 does not end on cylinder boundary:
     phys=(349, 65, 63) should be (349, 254, 63)

/dev/sdc4            91      263  1384614   83  Linux native
Partition 4 has different physical/logical beginnings (non-Linux?):
     phys=(350, 0, 1) logical=(90, 150, 1)
Partition 4 has different physical/logical endings:
     phys=(1015, 65, 63) logical=(262, 245, 63)
Partition 4 does not end on cylinder boundary:
     phys=(1015, 65, 63) should be (1015, 254, 63)

Note that all partitions, not just partition 1, seem to be
very confused.  Note that nowhere are we dealing with a cylinder
numbered > 1024, so this is not the traditional "BIOS can't handle
disks with lots of cylinders" problem.

This disk was healthy -- I've been booting off it for
2.5 years.  Is this a revision difference between Red Hat 5.1
fdisk and RH 5.2 fdisk?  Or is my disk really trashed?

de

--
.............................................................................
:De Clarke, Software Engineer                     UCO/Lick Observatory, UCSC:

:Web: www.ucolick.org | ones people hate, and the ones people don't use.--JO:

 
 
 

LILO 01 01 01 01 01 01....

Post by James Lothia » Sat, 20 Mar 1999 04:00:00


I had a similar problem to this a while ago. It seems that if
you have a machine with both ide and scsi disks, and you don't
tell lilo anything specific, it will assume that the ide disk
is your boot disk. I think that what you need to do is tell
lilo that your scsi disk is the disk that the bios is configured
to boot from, by including a "bios=<something>" line in your lilo.conf
file. If you don't do this, lilo will complain about your scsi disk
not being the 'first' disk, and you will get the 0101010 behaviour
when you try to boot.

The 'bios=' option isn't documented in the lilo man page. It is
however in the UserGuide.ps file  in /usr/doc/lilo-*, where this
problem is specifically described.

James


> Please send me mail if you have any ideas about this
> sad, sad story.

> I have a LILO boot problem...

> System:  P133 Advanced/EV board, NCR 53C825 PCI SCSI ctrlr,
> 96 MB RAM, SCSI boot disk sda, SCSI cdrom.  This system has
> been running *only* Linux for about 2.5 years, through a
> few new RH releases.  It had 32 MB until just last week :-)

> Recently I added 2 4GB IDE drives, which have been working
> fine for about 3 weeks.

> Then I added 2 32MB SIMMS to bring the memory complement
> up to 96MB.  But Linux only saw 64 MB.  I left that
> little problem for later...

> Then I got brave and tried installing an Evergreen Spectra
> 333 CPU upgrade.  Evergreen expects this upgrade to work
> on my Intel motherboard.

> The upgrade required replacement of the old AMI BIOS with
> MR BIOS.  Did that, and things seemed to go on working just
> fine.  Then I made the classic error of doing 2 things
> at once :-(

> I asked an Intel hw hacker about the memory issue and on his
> advice I modified lilo.conf to add append="mem=96m" and also
> shut down to install the new CPU.  When I rebooted with the
> new CPU I got as far as

>         LI_

> and that was it.

> I thought the Evergreen CPU didn't work.  So...
> put the old CPU back in and rebooted:

>         LI_

> and that was it :-(

> I tried booting RH install diskette in the thought that I
> could just reinstall the OS on the boot disk.  RH installer
> can't find the cdrom device on the SCSI bus.

> After a few iterations trying to get RH to see the CDROM,
> the boot behaviour changed to

>         [something, probably some portion of LILO, goes by
>          too fast to read] 01 01 01 01 01 01 01 01 01 01 01
>         01 01 01 01 01 01 01 01 01 01 01 01....

> you get the idea.  The 01 text is just pouring onto the
> screen at max scroll speed.

> So... ummm... I don't have a rescue diskette... I just
> upgraded to RH 5.1 recently and didn't get around to making
> it yet [yes, this is very embarrassing].

> Does anyone understand (a) what 01 01 01 01 means
>                  (b) why I can't see my CDROM any more
>                  (c) how I can ever get this crate booted again
> I tried typing in "linux single root=/dev/sda1" at the
> boot prompt on the RH install disk but that, obviously,
> didn't work... it does mount the ext2 fs on the SCSI disk,
> which gives me *hope* that the disk is still alive;  but
> I can't get a shell at all.

> Is there a simple hardware diag package on floppy disk that
> I could run to find out at least how much of my system
> is still alive?

> The NCR SCSI card is alive enough to report SCSI ID strings
> during initialization.  The rest of the system is alive
> enough that MR BIOS shows correct settings for memory
> size, CPU vendor and ID, etc.  There are really only 3
> symptoms:

> I can no longer boot from my SCSI boot disk
> I can no longer even SEE my SCSI CDROM from RH installer
>         so I can't reinstall

> and one "operator error"

> I don't have a rescue disk [stupid me!]

> Is my LILO problem due to the append statement I added
> or due to the death of my scsi controller, and how can
> I tell?  And is there any way to get this thing booted
> without its own special custom rescue disk?

> de

> --
> .............................................................................
> :De Clarke, Software Engineer                     UCO/Lick Observatory, UCSC:

> :Web: www.ucolick.org | ones people hate, and the ones people don't use.--JO:

 
 
 

LILO 01 01 01 01 01 01....

Post by Andries Brouw » Sat, 20 Mar 1999 04:00:00


: Please send me mail if you have any ideas about this
: sad, sad story.
...
: Then I added 2 32MB SIMMS to bring the memory complement
: up to 96MB.  But Linux only saw 64 MB.  I left that
: little problem for later...

: Then I got brave and tried installing an Evergreen Spectra
: 333 CPU upgrade.  Evergreen expects this upgrade to work
: on my Intel motherboard.

: The upgrade required replacement of the old AMI BIOS with
: MR BIOS.  Did that, and things seemed to go on working just
: fine.  Then I made the classic error of doing 2 things
: at once :-(

: I asked an Intel hw hacker about the memory issue and on his
: advice I modified lilo.conf to add append="mem=96m" and also
: shut down to install the new CPU.  When I rebooted with the
: new CPU I got as far as

:       LI_

: and that was it.

Reasonable enough: probably this new BIOS invents a new geometry
for your disk, and your lilo.conf did not have the `linear' keyword.

: I tried booting RH install diskette in the thought that I
: could just reinstall the OS on the boot disk.  RH installer
: can't find the cdrom device on the SCSI bus.

: After a few iterations trying to get RH to see the CDROM,
: the boot behaviour changed to

:       [something, probably some portion of LILO, goes by
:        too fast to read] 01 01 01 01 01 01 01 01 01 01 01
:       01 01 01 01 01 01 01 01 01 01 01 01....

: you get the idea.  The 01 text is just pouring onto the
: screen at max scroll speed.

So, now LILO can no longer find the second stage loader
and cannot access the disk where this second stage loader
is supposed to be.

: I tried typing in "linux single root=/dev/sda1" at the
: boot prompt on the RH install disk but that, obviously,
: didn't work... it does mount the ext2 fs on the SCSI disk,

Isnt that an enormous success? Then all is well.
It means that the SCSI controller is fine, and large parts
of the disk are fine.

: which gives me *hope* that the disk is still alive;  but
: I can't get a shell at all.

Use a rescue floppy (download it from somewhere).

: The NCR SCSI card is alive enough to report SCSI ID strings
: during initialization.  The rest of the system is alive
: enough that MR BIOS shows correct settings for memory
: size, CPU vendor and ID, etc.  There are really only 3
: symptoms:

: I can no longer boot from my SCSI boot disk

That part I can understand.

: I can no longer even SEE my SCSI CDROM from RH installer

Here I have no idea what is wrong.

: Is my LILO problem due to the append statement I added

No.

: or due to the death of my scsi controller

There is no reason to suppose the SCSI controller died, I think.

Does your new BIOS allow you to play with the geometry
of this SCSI disk? You might get it working again without
needing a rescue floppy.

 
 
 

LILO 01 01 01 01 01 01....

Post by Andries Brouw » Sat, 20 Mar 1999 04:00:00


: I had a similar problem with one SCSI and one IDE drive. If I ran LILO
: with the IDE drive connected, LILO would complain that the /dev/sda was
: not the first drive, but would still run (even if my BIOS said to use
: SCSI as boot drive instead of IDE [this is windoze stuff]), except that
: when I reboot, I get the LI error...

: My solution is that I need to unplug the IDE drive (yes unplug, just
: telling the BIOS that the 'C' drive is 'none' is not good enough, Linux
: still finds 'hda') and then boot with a rescue disk and run:

: mount /dev/sdaX /mnt ; lilo -r /mnt ; umount /mnt

: After that, I can shutdown and reconnect the IDE drive and then
: everybody's happy...

: I have not been able to find references to this 'SCSI-IDE-LILO' problem
: (haven't looked that hard)

Ah, how primitive...

There is no need to go through such contortions.
Put a disk section in /etc/lilo.conf.
Quoting a manual page:

    disk=device-name
          Defines  non-standard  parameters for the specified
          disk.  See section "Disk geometry" of user.tex  for
          details.   Especially  useful  is  a `disk' section
          like

          disk=/dev/sda
              bios=0x80

          which tells LILO that your system boots from a SCSI
          disk, even though also IDE disks may be present.

 
 
 

LILO 01 01 01 01 01 01....

Post by Andries Brouw » Sat, 20 Mar 1999 04:00:00


: I just spent a week trying to sort mine out, I was getting all these 01 01
: 01 just constantly appearing on the screen, I can't install anything on that
: hard disk now tried deleting all partitions re-formatting it, and putting NT
: back on it and it still comes up with the 01's. So basically now I have a
: hard disk that'll never boot again. All because LILO tried to stick itself
: in the MBR.

: Oh and at the same time it managed to damage the hard drive I used to
: install Linux from... Looks like I'll have to download it all again, great
: eh?

Not so pessimistic! Such things are always very simple.
For example, if you have a Linux system, say
        dd if=/dev/zero of=/dev/whatever bs=512 count=1
to wipe the MBR record clean.

If you cannot do this from Linux, say
        FDISK /MBR
from a DOS floppy with FDISK.

By the way, probably you didnt damage anything.

 
 
 

LILO 01 01 01 01 01 01....

Post by Andries Brouw » Sat, 20 Mar 1999 04:00:00


: I had a similar problem to this a while ago. It seems that if
: you have a machine with both ide and scsi disks, and you don't
: tell lilo anything specific, it will assume that the ide disk
: is your boot disk. I think that what you need to do is tell
: lilo that your scsi disk is the disk that the bios is configured
: to boot from, by including a "bios=<something>" line in your lilo.conf
: file. If you don't do this, lilo will complain about your scsi disk
: not being the 'first' disk, and you will get the 0101010 behaviour
: when you try to boot.

: The 'bios=' option isn't documented in the lilo man page. It is
: however in the UserGuide.ps file  in /usr/doc/lilo-*, where this
: problem is specifically described.

Yes, precisely.
But the 'bios=' option is now documented in the lilo.conf man page.
It will be in man-pages-1.23.

[Now I happen to come across this remark, but it is not
forbidden to write to the maintainer of the man pages
when something is wrong or absent. You might cause things

 
 
 

LILO 01 01 01 01 01 01....

Post by Andries Brouw » Thu, 25 Mar 1999 04:00:00


: I've been doggedly pursuing my LILO problem.  The next logical
: step was to check out the hardware.

: I put the SCSI boot disk on another machine's scsi bus (known
: to work) and took a look at it with fdisk:  odd results!

: Disk /dev/sdc: 255 heads, 63 sectors, 263 cylinders
: Units = cylinders of 16065 * 512 bytes

:    Device Boot    Start      End   Blocks   Id  System

: /dev/sdc1   *         1        9    66496+  83  Linux native
: Partition 1 has different physical/logical endings:
:      phys=(31, 65, 63) logical=(8, 71, 63)
: Partition 1 does not end on cylinder boundary:
:      phys=(31, 65, 63) should be (31, 254, 63)

This is a good disk, that was written on a system that used
a geometry with 66 heads and 63 sectors/track, and now you read
it on a SCSI controller that uses 255 heads and 63 sectors/track.
(This has no connection whatsoever with the hardware, they are
just random numbers to please DOS which needs such numbers.)

: Note that all partitions, not just partition 1, seem to be
: very confused.  Note that nowhere are we dealing with a cylinder
: numbered > 1024, so this is not the traditional "BIOS can't handle
: disks with lots of cylinders" problem.

: This disk was healthy -- I've been booting off it for
: 2.5 years.  Is this a revision difference between Red Hat 5.1
: fdisk and RH 5.2 fdisk?  Or is my disk really trashed?

Nothing wrong with your disk.
No problem in sight anywhere.
This disk will work perfectly with Linux.
(This geometry business is only a matter for DOS and BIOS.)

 
 
 

LILO 01 01 01 01 01 01....

Post by Mark Aldridg » Sat, 03 Apr 1999 04:00:00


From the heading,as it is not clear in the note below, I guess you are getting a
series of 01's scrolling up the screen when you try and boot, as I have had the
same problem a few times.

The way I have cured this ,without any loss of data, is to use FDISK. You will
need to boot from a DOS floppy, and at the command prompt enter FDISK /MBR. This
restores the master boot record to the disk. Remove the floppy reboot, and
everything should be ok.

I have used this about three times, in my attempts to install LILO (to boot
windows98 from an IDE disk, and Linux from a SCSI disk). Twice it worked just
using the command above, the third time I had to use the DOS command SYS C: ,
after using the FDISK command above, once I had booted from a Windows98 startup
floppy (use a floppy created using the same windows version as installed on the
hard drive, to avoid version conflicts).

Hope this helps

MA


> I've been doggedly pursuing my LILO problem.  The next logical
> step was to check out the hardware.

> I put the SCSI boot disk on another machine's scsi bus (known
> to work) and took a look at it with fdisk:  odd results!

> Disk /dev/sdc: 255 heads, 63 sectors, 263 cylinders
> Units = cylinders of 16065 * 512 bytes

>    Device Boot    Start      End   Blocks   Id  System

> /dev/sdc1   *         1        9    66496+  83  Linux native
> Partition 1 has different physical/logical endings:
>      phys=(31, 65, 63) logical=(8, 71, 63)
> Partition 1 does not end on cylinder boundary:
>      phys=(31, 65, 63) should be (31, 254, 63)

> /dev/sdc2             9       26   135135   82  Linux swap
> Partition 2 has different physical/logical beginnings (non-Linux?):
>      phys=(32, 0, 1) logical=(8, 72, 1)
> Partition 2 has different physical/logical endings:
>      phys=(96, 65, 63) logical=(25, 26, 63)
> Partition 2 does not end on cylinder boundary:
>      phys=(96, 65, 63) should be (96, 254, 63)

> /dev/sdc3            26       91   525987   83  Linux native
> Partition 3 has different physical/logical beginnings (non-Linux?):
>      phys=(97, 0, 1) logical=(25, 27, 1)
> Partition 3 has different physical/logical endings:
>      phys=(349, 65, 63) logical=(90, 149, 63)
> Partition 3 does not end on cylinder boundary:
>      phys=(349, 65, 63) should be (349, 254, 63)

> /dev/sdc4            91      263  1384614   83  Linux native
> Partition 4 has different physical/logical beginnings (non-Linux?):
>      phys=(350, 0, 1) logical=(90, 150, 1)
> Partition 4 has different physical/logical endings:
>      phys=(1015, 65, 63) logical=(262, 245, 63)
> Partition 4 does not end on cylinder boundary:
>      phys=(1015, 65, 63) should be (1015, 254, 63)

> Note that all partitions, not just partition 1, seem to be
> very confused.  Note that nowhere are we dealing with a cylinder
> numbered > 1024, so this is not the traditional "BIOS can't handle
> disks with lots of cylinders" problem.

> This disk was healthy -- I've been booting off it for
> 2.5 years.  Is this a revision difference between Red Hat 5.1
> fdisk and RH 5.2 fdisk?  Or is my disk really trashed?

> de

> --
> .............................................................................
> :De Clarke, Software Engineer                     UCO/Lick Observatory, UCSC:

> :Web: www.ucolick.org | ones people hate, and the ones people don't use.--JO:

 
 
 

1. lilo li 01 01 01 01 01 01 01 01 solution

please, no flames, I am just saying what worked for me.

I went from Caldera 1.3 and Idiots Guide Book (steep learning curve there),
to Caldera 2.2 and Using Caldera Open Linux 2.2 and then a Cheapbytes 2.3
Caldera disk.  System was working the way I wanted it to.  I was having fun
with a dual boot that saw less and less of Windows, days without.  Months
pass

I lent my computer to mom for two months as I moved up to Canada, I unhooked
hdb so she would not have to be dual booting, let Doze98 overwrite the mbr,
but hey, I got a floppy with LILO.

Upon bringing computer back, and reinstalling hdb, I kept running into the
li 01 01 01 01 01 01 01 01 01 due to a probable geometry mismatch.  fdisked
mbr, reinstalled, yanked out sound and network cards, disabled com irq's,
Tried with Caldera 2.3, 2.2, phat linux which is an older mandrake that will
run in fat32, swapped drives back and forth between primary and secondary
(made sure jumpers were correct on hdd's) nothing works.  I keep getting li
and then a cascade of 01 01 01 01 01 01 01 01 01 sweeping down the screen.
I was going to try nuni - finally I pulled out an old FreeBSD 3.2 cd-rom
that I had played with on a 486 - it went on sweet as silk.   Dual booted
flawlessly.

Downloaded boot floppy disks of FreeBSD 3.4 and proceeded to ftp  /bin and
/doc from a mirror.  This is on a cable modem mind you, I was not expecting
to be able to do that.  My CD-rom for 3.4 came the next day and I finished
it up.  I don't have mail working yet, but lynx is working quite well as is
ftp.

I intend to reinstall Linux one day in the future.  I have a new Mandrake to
try out.  But it is going to share a swap file with FreeBSD and use the
FreeBSD loader.

Shalom,  Mark Rogness

2. hostadaptor installation

3. 01 01 01 01 01 01 01 01

4. SunISDN 1.0.4 and the Sparcbook3

5. SuSE 7: 01 01 01 01 01 01 01 ...

6. Linksys Router via Telnet

7. lilo won't start, sez: L01 01 01 01 01 01 01 01 01 01 01 01 01

8. better way ?

9. Linux installation problems - 01 01 01 01 01 01 ...

10. L01 01 01 01 01 01 01...

11. Trying to boot, just get "L 01 01 01 01 01 01 01 01 01 01 01..."

12. 01 01 01 01 01 01 01 01 01 01 from lilo? (on RedHat 4.0 & 4.1)

13. L 01 01 01 01 01