Error
messages from LILO
The information in this section has been copied from the LiLo Manual, so
none of this is my work. I plan to try editing it a little for easier reading and then add some
hyperlinks for your convenience.
You will find your own copy of the
LiLo manual in your Ubuntu file system if you have LiLo installed,
Get the Lilo Manual.
All parts of LILO display some messages that can be used to diagnose problems.
Boot loader messages
The boot loader generates three types of messages:
- progress and error messages while it is loading,
- messages indicating disk access errors,
- and error messages in response to invalid command-line input.
Since messages of the latter type are usually self-explanatory, only the two other categories are explained.
LILO start messages
When LILO loads itself, it displays the word "LILO".
Each letter is printed before or after performing some specific action.
If LILO fails at some point, the letters printed so far can be used to
identify the problem. This is described in more detail in the technical
overview.
Note that some hex digits may be inserted after the first "L"
if a transient disk problem occurs. Unless LILO stops at that point,
generating an endless stream of error codes, such hex digits do not
indicate a severe problem.
(<nothing>)
No part of LILO has been loaded. LILO either isn't installed or
the partition on which its boot sector is located isn't active.
L <error> ...
The first stage boot loader has been loaded and started, but it can't
load the second stage boot loader. The two-digit error codes
indicate the type of problem. (See also section "Disk
error codes".) This condition usually indicates a media failure or
a geometry mismatch (e.g. bad disk parameters, see section "Disk
geometry").
LI
The first stage boot loader was able to load the second stage
boot loader, but has failed to execute it. This can either be
caused by a geometry mismatch or by moving /boot/boot.b without
running the map installer.
LIL The second
stage boot loader has been started, but it can't load the
descriptor table from the map file. This is typically caused by
a media failure or by a geometry mismatch.
LIL? The second
stage boot loader has been loaded at an incorrect address. This is
typically caused by a subtle geometry mismatch or by moving
/boot/boot.b without running the map installer.
LIL- The
descriptor table is corrupt. This can either be caused by
a geometry mismatch or by moving /boot/map without running the
map installer.
LILO All parts of LILO have been successfully loaded. |
Map installer warnings and errors
Most messages of the map installer (/sbin/lilo) should be self-explanatory.
Some messages that indicate common errors are listed below.
They are grouped into fatal errors and warnings (non-fatal errors).
Fatal errors
Boot sector of <device_name> doesn't have a boot signature
Boot sector of <device_name> doesn't have a LILO signature
The sector from which LILO should be uninstalled doesn't appear to be a LILO boot sector.
Can't put the boot sector on logical partition <number>
An attempt has been made to put LILO's boot sector on the current
root file system partition which is on a logical partition. This
usually doesn't have the desired effect, because common MBRs can
only boot primary partitions. This check can be bypassed by
explicitly specifying the boot partition with
the -b option or by setting the configuration variable
BOOT.
Checksum error
The descriptor table of the map file has an invalid checksum. Refresh the map file _immediately_ !
Device 0x<number>: Configured as inaccessible.
There
is a DISK section entry indicating that the device is inaccessible
from the BIOS. You should check carefully that all files LILO
tries to access when booting are on the right device.
Device 0x<number>: Got bad geometry <sec>/<hd>/<cyl>
The device driver for your SCSI controller does not support
geometry detection. You have to specify the geometry explicitly
(see section "Disk geometry").
Device 0x<number>: Invalid partition table, entry <number>
The 3D and linear addresses of the first sector of the
specified partition don't correspond. This is typically caused by
partitioning a disk with a program that doesn't align partitions
to tracks and later using PC/MS-DOS or OS/2 on that disk. LILO can
attempt to correct the problem, see "General per-image options".
Device 0x<number>: Partition type 0x<number> does not seem suitable for a LILO boot sector
The
location where the LILO boot sector should be placed does not seem to
be suitable for that. (See also also section "Disk organization").
You should either adjust the partition type to reflect the actual
use or put the boot sector on a different partition. This
consistency check only yields a warning (i.e. LILO continues) if
the option IGNORE-TABLE is set.
<device_name> is not a valid partition device
The specified device is either not a device at all, a whole disk, or a
partition on a different disk than the one in whose section its
entry appears.
<device_name> is not a whole disk device
Only the geometry of whole disks (e.g. /dev/hda, /dev/sdb, etc.) can be redefined when using DISK sections.
DISKTAB and DISK are mutually exclusive
You cannot use a disktab file and disk geometry definitions in
the configuration file at the same time. Maybe /etc/disktab
was accidentally used, because that's the default for
backward-compatibility. You should delete /etc/disktab after
completing the transition to DISK sections.
Duplicate entry in partition table
A partition table entry appears twice. The partition table has to be fixed with fdisk.
Duplicate geometry definition for <device_name>
A disk or partition geometry definition entry for the same
device appears twice in the configuration file. Note that you
mustn't write a partition section for the whole disk - its start
sector is always the first sector of the disk.
First sector of <device> doesn't have a valid boot signature
The first sector of the specified device does not appear to be a
valid boot sector. You might have confused the device name.*
* Because different partition programs may display the partitions in
a different order, it is possible that what you think is your
first partition isn't /dev/hda1, etc. A good method to verify the
content of a partition is to try to mount it.
geo_comp_addr: Cylinder <number> beyond end of media (<number>)
A file block appears to be located beyond the last cylinder of
the disk. This probably indicates an error in the disk geometry
specification (see section "Disk geometry") or a file system
corruption.
geo_comp_addr: Cylinder number is too big (<number> > 1023)
Blocks of a file are located beyond the 1024th cylinder of a hard disk.
LILO can't access such files, because the BIOS limits cylinder numbers to the range 0...1023.
Try
moving the file to a different place, preferably a partition that is
entirely within the first 1024 cylinders of the disk.
Hole found in map file (<location>)
The map installer is confused about the disk organization. Please report this error.
<item> doesn't have a valid LILO signature
The specified item has been located, but is not part of LILO.
<item> has an invalid stage code (<number>)
The specified item has probably been corrupted. Try re-building LILO.
<item> is version <number>. Expecting version <number>.
The specified entity is either too old or too new. Make sure all
parts of LILO (map installer, boot loaders and chain loaders) are
from the same distribution. **
** The expected version number may be different from the version
number of the LILO package, because file version numbers are only
increased when the file formats change.
Kernel <name> is too big
The kernel image (without the setup code) is bigger than 512
kbytes (or 448 kbytes, if built with LARGE_EDBA ). LILO
would overwrite itself when trying to load such a kernel.
This
limitation only applies to old kernels which are loaded below
0x10000 (e.g. "Image" or "zImage"). Try building the kernel with
"bzImage".
If this is undesirable for some reason, try
removing some unused drivers and compiling the kernel again. This
error may also occur if the kernel image is damaged or if it
contains trailing "junk", e.g. as the result of copying an entire
boot floppy to the hard disk.
LOCK and FALLBACK are mutually exclusive
Since LOCK and FALLBACK both change the default command line, they can't be reasonably used together.
Map <path> is not a regular file.
This is probably the result of an attempt to omit writing a map
file, e.g. with -m /dev/null . The -t option
should be used to accomplish this.
Must specify SECTORS and HEADS together
It is assumed that disks with a "strange" number of sectors will also
have a "strange" number of heads. Therefore, it's all or nothing.
No geometry variables allowed if INACCESSIBLE
If a device is configured as INACCESSIBLE (see section "Specifying
the geometry"), its DISK section must not contain any geometry
variables.
No image <image> is defined
The command line specified either with the -R option or
with FALLBACK does not contain the name of a valid image. Note
that optional images which have not been included in the map file
are not considered as valid.
Partition entry not found
The partition from which an other operating system should be
booted isn't listed in the specified partition table. This either
means that an incorrect partition table has been specified or that
you're trying to boot from a logical partition. The latter usually
doesn't work. You can bypass this check by omitting the partition
table specification (e.g. omitting the variable TABLE).
Single-key clash: "<name>" vs. "<name>"
The specified image labels or aliases conflict because one of them
is a single character and has the SINGLE-KEY option set, and the
other name begins with that character.
Sorry, don't know how to handle device <number>
LILO uses files that are located on a device for which there is
no easy way to determine the disk geometry. Such devices have to
be explicitly described, see section "Disk geometry".
This LILO is compiled READONLY and doesn't support ...
If LILO is not allowed to write to the disk at boot time (see
section "Build-time configuration"), options like LOCK and
FALLBACK are unavailable.
This LILO is compiled without REWRITE_TABLE and doesn't support ...
If LILO is not allowed to rewrite partition tables at boot time
(see section "Partition table manipulation"), options like
ACTIVATE and SET (in a CHANGE section) are unavailable. You may
also get this error if LILO is compiled with READONLY
enabled.
Timestamp in boot sector of <device> differs from date of <file>
The backup copy of the boot sector does not appear to be an
ancestor of the current boot sector. If you are absolutely sure
that the boot sector is indeed correct, you can bypass this check
by using -U instead of -u .
Trying to map files from unnamed device 0x<number> (NFS ?)
This is probably the same problem as described below, only with the root file system residing on NFS.
Trying to map files from your RAM disk.
Please
check -r option or ROOT environment variable. Most likely,
you or some installation script is trying to invoke LILO in a way
that some of the files is has to access reside on the RAM disk.
Normally, the ROOT environment variable should be set to
the mount point of the effective root device if installing LILO
with a different root directory. See also sections "Create or
update map" and "Normal first-time installation".
VGA mode presetting is not supported by your kernel.
Your kernel sources appear to be very old ('93 ?). LILO may work on your system if you remove the VGA option.
write <item>: <error_reason>
The disk is probably full or mounted read-only.
Warnings
Messages labeled with "Warning" can be turned off with the NOWARN option.
FIGETBSZ <file_name>: < error_reason>
The map installer is unable to determine the block size of a file
system. It assumes a block size of two sectors (1kB).
Ignoring entry '<variable_name>'
The command-line option corresponding to the specified variable is
set. Therefore, the configuration file entry is ignored.
Setting DELAY to 20 (2 seconds)
Because accidentally booting the wrong kernel or operating system may
be very inconvenient on systems that are not run from a local display,
the minimum delay is two seconds if the SERIAL variable is set.
(temp) <item>: <error_reason>
Deleting a temporary file has failed for the specified reason.
Warning: BIOS drive 0x<number> may not be accessible
Because most BIOS versions only support two floppies and two hard
disks, files located on additional disks may be inaccessible. This
warning indicates that some kernels or even the whole system may be
unbootable.
Warning: COMPACT may conflict with LINEAR on some systems
Please see section "Other problems" for a description of this problem.
Warning: <config_file> should be owned by root
In order to prevent users from compromising system integrity, the
configuration file should be owned by root and write access for all
other users should be disabled.
Warning: <config_file> should be readable only for root if using PASSWORD
Users should not be allowed to read the configuration file when using
the PASSWORD option, because then, it contains unencrypted passwords.
Warning: <config_file> should be writable only for root
See " Warning: <config_file> should be owned by root ".
Warning: device 0x<number> exceeds 1024 cylinder limit
A disk or partition exceeds the 1024 cylinder limit imposed by the BIOS.
This may result in a fatal error in the current installation run
or in later installation runs.
See " geo_comp_addr: Cylinder number is too big (<number> > 1023) " for details.
Warning: <device> is not on the first disk
The specified partition is probably not on the first disk. LILO's boot
sector can only be booted from the first disk unless some special boot
manager is used.
WARNING: The system is unbootable !
One of the last installation steps has failed. This warning is
typically followed by a fatal error describing the problem.
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".
0x02 "Address mark not found".
This usually indicates a media problem. Try again several times.
0x03 "Write-protected disk".
This should only occur on write operations.
0x04 "Sector not found".
This typically indicates a geometry mismatch.
If
you're booting a raw-written disk image, verify whether it
was created for disks with the same geometry as the one you're
using.
If you're booting from a SCSI disk or a large IDE disk, you
should check whether LILO has obtained correct geometry data from the
kernel or whether the geometry definition corresponds to the real disk
geometry.
(See section "Disk geometry".) Removing COMPACT may help too. So may adding LBA32 or LINEAR.
0x06 "Change line active".
This should be a transient error. Try booting a second time.
0x07 "Invalid initialization".
The
BIOS failed to properly initialize the disk controller. You should
control the BIOS setup parameters. A warm boot might help too.
0x08 "DMA overrun".
This shouldn't happen. Try booting again.
0x09 "DMA attempt across 64k boundary".
This
shouldn't happen, but may inicate a disk geometry mis-match. Try
omitting the COMPACT option. You may need to specify the disk
geometry yourself.
0x0C "Invalid media".
This shouldn't happen and might be caused by a media error. Try booting again.
0x10 "CRC error".
A
media error has been detected. Try booting several times, running
the map installer a second time (to put the map file at some other
physical location or to write "good data" over the bad spot), mapping
out the bad sectors/tracks and, if all else fails, replacing the
media.
0x11 "ECC correction successful".
A
read error occurred, but was corrected. LILO does not recognize
this condition and aborts the load process anyway. A second load
attempt should succeed.
0x20 "Controller error".
This shouldn't happen.
0x40 "Seek failure".
This might be a media problem. Try booting again.
0x80 "Disk timeout".
The
disk or the drive isn't ready. Either the media is bad or the disk
isn't spinning. If you're booting from a floppy, you might not
have closed the drive door. Otherwise, trying to boot again might
help.
0xBB "BIOS error".
This
shouldn't happen. Try booting again. If the problem persists,
removing the COMPACT option or adding/removing LINEAR or LBA32
might help.
If the error occurred during a write operation, the error code (two hex digits) is prefixed with a "W".
Although
write errors don't affect the boot process, they might indicate a
severe problem, because they usually imply that LILO has tried to write
to an invalid location.
If spurious write errors occur on a
system, it might be a good idea to configure LILO to run read-only (see
section "Build-time configuration").
Generally, invalid geometry and attempts to use more than two disks
without a very modern BIOS may yield misleading error codes. Please
check carefully if /sbin/lilo doesn't emit any warnings. Then try using
the LINEAR or LBA32 option (see section "Global options").
Other problems
This section contains a collection of less common problems that have been
observed. See also section "Installation of other operating systems" for
general remarks on using LILO with other operating systems. Some of the
problems are obscure and so are the work-arounds.
- If LILO doesn't go away even if you erase its files, format your Linux
partition, etc., you've probably installed LILO as your MBR and you've
forgotten to deinstall it before deleting its files. See section "LILO
de-installation" for what you can do now.
- For yet unknown reasons, LILO may fail on some systems with AMI BIOS if
the "Hard Disk Type 47 RAM area" is set to "0:300" instead of "DOS 1K".
- Some disk controller BIOSes perform disk geometry/address translations
that are incompatible with the way the device's geometry is seen from
Linux, i.e. without going through the BIOS. Particularly, large IDE
disks and some PCI SCSI controllers appear to have this problem. In
such cases, either the translated geometry has to be specified in a
DISK section or the sector address translation can be deferred by using
the LINEAR option. In a setup where floppies are not normally used for
booting, the LINEAR approach should be preferred, because this avoids
the risk of specifying incorrect numbers.
- OS/2 is said to be bootable from a logical partition with LILO acting
as the primary boot selector if LILO is installed on the MBR, the OS/2
BootManager is on an active primary partition and LILO boots
BootManager. Putting LILO on an extended partition instead is said to
crash the OS/2 FDISK in this scenario.
Note that booting LILO from BootManager (so BootManager is the primary
selector) or booting OS/2 directly from a primary partition (without
BootManager) should generally work. See also section "Installation of
other operating systems".
- Windows NT is reported to be bootable with LILO when LILO acts as the
MBR and the Windows NT boot loader is on the DOS partition. However,
NT's disk manager complains about LILO's MBR when trying to edit the
partition table.
- Some PC UNIX systems (SCO and Unixware have been reported to exhibit
this problem) depend on their partition being active. See section
"Partition table manipulation" for how this can be accomplished.
- Future Domain TMC-1680 adapters with the BIOS versions 3.4 and 3.5
assign BIOS device numbers in the wrong order, e.g. on a two-disk
system, /dev/sda becomes 0x81 and /dev/sdb becomes 0x80 . This can
be fixed with the following DISK section:
disk=/dev/sda bios=0x81 disk=/dev/sdb bios=0x80
Note that this is only valid for a two-disk system. In three-disk
systems, /dev/sdc would become 0x80 , etc. Also, single-disk systems
don't have this problem (and the "fix" would break them).
- Some BIOSes don't properly recognize disks with an unusual partition
table (e.g. without any partition marked active) and refuse to boot
from them. This can also affect the second hard disk and the problem
may only occur if the system is booted in a particular way (e.g. only
after a cold boot).
- On some systems, using LINEAR and COMPACT or LBA32 and COMPACT together
leads to a boot failure. The exact circumstances under which this
happens are still unknown.
- If the kernel crashes after booting on a multi-processor system, LILO
may have overwritten data structures set up by the BIOS. Try the option
LARGE_EBDA in this case.
|
