======================================================================
EXPLANATION OF HPFS386 INTERNAL ERRORS
======================================================================
Introduction
------------
When the LAN Manager version 2.00 installable file system (HPFS386)
encounters unrecoverable errors, it stops the system and displays the
error on the screen using the following format:
Internal processing error
At location #01F0:0892
HPFS Fatal Internal Error #xx
The following is a list of the fatal internal errors HPFS386 may
display, along with a brief explanation and possible courses of action
for correcting the problem:
Internal Error #0
-----------------
Explanation The disk driver routine returned to an HPFS interrupt
handler an invalid command.
Action If you are using standard devices (for example, hard
disks that have a proven history under OS/2), the most
likely cause is corrupted memory (perhaps caused by some
other errant driver). If you are using an unusual
device, the problem may be in the device driver supplied
by the device manufacturer.
Internal Error #1
-----------------
Explanation The file system was unable to allocate a new directory
block. Allocation (and deallocation) of directory blocks
may be caused by something as simple as renaming a file.
When a new directory block is needed, HPFS first tries
to allocate it from the band on the disk dedicated to
directory blocks. If that fails (either because no more
sectors are available or an error occurred accessing the
disk), HPFS attempts to allocate the directory block
from another intelligent location on the disk. This
error is issued only when all attempts to allocate a
directory block have failed.
Action This error can be caused by a full disk. If the disk is
not full, the error may be symptomatic of a disk
failure. Run Chkdsk /F against the disk to reset the
hotfix sector and correct any problems that may exist.
If the problem still occurs (or if Chkdsk cannot
complete), try replacing the disk drive and related
hardware (for example, the controller).
Internal Error #3
-----------------
Explanation A failure occurred reading the hotfix sector. During
the process of a hot fix (that is, the dynamic
replacement of a bad sector with a good sector in the
allocation of a file), HPFS reads the one sector bitmap
describing which sectors have been reserved for dynamic
replacement (that is, hotfix sectors). An error was
returned to HPFS when it attempted to read this hotfix
sector.
Action This error can occur only after a failure to read or
write another sector on the disk; therefore, it is
likely caused by disk drive or hardware problems. HPFS
cannot recover if the sector containing the hotfix
bitmap goes bad. Run Chkdsk /F to try to recover the
volume. If the problem still occurs (or if Chkdsk
cannot complete), try replacing the disk drive and
related hardware (for example, the controller).
Internal Error #4
-----------------
Explanation A failure occurred writing the hotfix sector. See the
previous explanation for additional details.
Action See previous action for recommended steps.
Internal Error #5
-----------------
Explanation Failure to allocate storage for the DBCS portion of the
codepage. During initialization, HPFS fetches the
codepage(s) in use by OS/2 and builds a translation
table. The first part of the translation table is placed
into preallocated storage. However, storage is allocated
to hold the DBCS portion (since it may not exist). HPFS
failed to obtain the storage necessary to hold this
portion of the codepage.
Action This is a general system error and does not indicate any
problems with the disk drive or its related hardware. If
the problem can be is re-created, you should notify a
Microsoft Product Support representative for the
appropriate product (for example, OS/2, LAN Manager).
Internal Error #7
-----------------
Explanation A hotfix attempt was made during HPFS shutdown. During
normal processing, HPFS attempts to recover from bad
sectors by hotfixing the data to a good sector (see
explanation under error number 3). However, during
shutdown, HPFS cannot employ hotfixes to recover. This
error indicates that a bad sector was discovered during
shutdown and HPFS cannot recover.
Action As with other errors that can indicate disk drive or
hardware problems, run Chkdsk /F against the drive. If
the problem still occurs (or if Chkdsk cannot complete),
try replacing the disk drive and related hardware (for
example, the controller).
Internal Error #8
-----------------
Explanation The main file system driver (FSD) was unable to open a
file that had been opened by the minimal function file
system driver (mini-FSD) used during boot time. When an
installable file system installs, control is given to a
mini-FSD. This mini-FSD is used to mount volumes and
open files used during the early stages of the boot
process. In the later stages, the full-function FSD is
given control. To maintain the same image as the
mini-FSD, all volumes mounted or files opened by the
mini-FSD must be mounted or opened by the full function
FSD. This error indicates that a failure occurred while
opening a file that had been opened by the mini-FSD.
Action This error does not normally imply a failure with the
disk drive or its related hardware. However, you may
want to try running Chkdsk /F against the drive. If
Chkdsk cannot complete, try replacing the disk drive or
its related hardware (for example, the controller). If
Chkdsk does complete and the problem recurs, this may
indicate a problem in either the device driver or HPFS.
Contact a Microsoft Product Support representative for
the appropriate product (for example, OS/2, LAN Manager)
for additional assistance.
Internal Error #9
-----------------
Explanation HPFS has exhausted its supply of hotfix sectors. Each
time HPFS detects a bad sector, it reassociates the data
with a good sector (that is, it hotfixes the problem). A
fixed quantity of hotfix sectors are preallocated when
the device is formatted and a bitmap is created that
records which hotfix sectors are available. This error
indicates that all of the hotfix sectors that were
preallocated have been used.
Action One of the tasks performed by Chkdsk /F on an HPFS
formatted device is to move data out of the special
hotfix sectors and into a free good sector elsewhere on
the device. You should run Chkdsk /F against the device
to empty the hotfix sectors and make them available
again to HPFS. However, you may also want to have the
disk drive and its related components checked. Running
out of hotfix sectors indicates that the device is
experiencing a larger than average number of failures
during reads or writes.
As a final note, these errors are not specific to the LAN Manager
version of HPFS (HPFS386). The information here should also apply to
volumes formatted with the standard OS/2 HPFS. Also, it is a good
rule of thumb to run Chkdsk /F before reaccessing any drive that has
experienced an error. If the data on the device is critical, you may
also want to have the device and its related hardware checked.
Finally, the best advice is to always keep critical data backed up on
secondary devices, since you can never tell when a problem may occur.