Input/output error
Adrian Wontroba
aw1 at stade.co.uk
Mon Nov 25 03:40:59 GMT 2002
--dDRMvlgZJXvWKvBx
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
On Sun, Nov 24, 2002 at 11:00:32PM +0000, Paul Civati wrote:
> ISTR looking at bad144 information at the time, there seemed to be a few
> references to it lurking in some man pages. ;)
One of my boxes is quite old, started out with 2.something, and has a
lot of ancient stuff hanging around it (one if the disadvantages of
source upgrades). Attached is a bad144 man page.
--
Adrian Wontroba
--dDRMvlgZJXvWKvBx
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename=bad144
BAD144(8) FreeBSD System Manager's Manual BAD144(8)
NAME
bad144 - read/write dec standard 144 bad sector information
SYNOPSIS
bad144 [-c] [-f] [-v] disk [sno [bad ...]]
bad144 -a [-c] [-f] [-v] disk [bad ...]
bad144 [-s] [-v] disk
DESCRIPTION
Bad144 can be used to inspect the information stored on a disk that is
used by the disk drivers to implement bad sector forwarding.
Available options:
-a The argument list consists of new bad sectors to be added to an
existing list. The new sectors are sorted into the list, which
must have been in order. Replacement sectors are moved to accom-
modate the additions; the new replacement sectors are cleared.
-c Forces an attempt to copy the old sector to the replacement, and
may be useful when replacing an unreliable sector.
-f For a RP06, RM03, RM05, Fujitsu Eagle, or SMD disk on a Massbus,
the -f option may be used to mark the new bad sectors as ``bad''
by reformatting them as unusable sectors. This option is
required unless the sectors have already been marked bad, or the
system will not be notified that it should use the replacement
sector. This option may be used while running multiuser; it is
no longer necessary to perform format operations while running
single-user.
-s The entire disk is scanned for bad blocks.
-v The entire process is described as it happens in gory detail if
-v (verbose) is given.
The format of the information is specified by DEC standard 144, as fol-
lows. The bad sector information is located in the first 5 even numbered
sectors of the last track of the disk pack. There are five identical
copies of the information, described by the dkbad structure.
Replacement sectors are allocated starting with the first sector before
the bad sector information and working backwards towards the beginning of
the disk. A maximum of 126 bad sectors are supported. The position of
the bad sector in the bad sector table determines the replacement sector
to which it corresponds. The bad sectors must be listed in ascending
order.
The bad sector information and replacement sectors are conventionally
only accessible through the ``c'' file system partition of the disk. If
that partition is used for a file system, the user is responsible for
making sure that it does not overlap the bad sector information or any
replacement sectors. Thus, one track plus 126 sectors must be reserved
to allow use of all of the possible bad sector replacements.
The bad sector structure is as follows:
struct dkbad {
long bt_csn; /* cartridge serial number */
u_short bt_mbz; /* unused; should be 0 */
u_short bt_flag; /* -1 => alignment cartridge */
struct bt_bad {
u_short bt_cyl; /* bad sector cylinder number */
u_short bt_trksec; /* track and sector number */
} bt_bad[126];
};
Unused slots in the bt_bad array are filled with all bits set, a puta-
tively illegal value.
Bad144 is invoked by giving a device name (e.g. hk0, hp1, etc.). With no
optional arguments it reads the first sector of the last track of the
corresponding disk and prints out the bad sector information. It issues
a warning if the bad sectors are out of order. Bad144 may also be
invoked with a serial number for the pack and a list of bad sectors. It
will write the supplied information into all copies of the bad-sector
file, replacing any previous information. Note, however, that bad144
does not arrange for the specified sectors to be marked bad in this case.
This procedure should only be used to restore known bad sector informa-
tion which was destroyed.
It is no longer necessary to reboot to allow the kernel to reread the
bad-sector table from the drive.
SEE ALSO
badsect(8)
BUGS
It should be possible to format disks on-line under UNIX.
It should be possible to mark bad sectors on drives of all type.
On an 11/750, the standard bootstrap drivers used to boot the system do
not understand bad sectors, handle ECC errors, or the special SSE (skip
sector) errors of RM80-type disks. This means that none of these errors
can occur when reading the file /kernel to boot. Sectors 0-15 of the
disk drive must also not have any of these errors.
The drivers which write a system core image on disk after a crash do not
handle errors; thus the crash dump area must be free of errors and bad
sectors.
HISTORY
The bad144 command appeared in 4.1BSD.
4th Berkeley Distribution May 13, 1995 4th Berkeley Distribution
--dDRMvlgZJXvWKvBx--
More information about the Ukfreebsd
mailing list