Eh?  fsck should be run on an UNmounted filesystem,
using its raw (unbuffered) device (I don't think Linux
offers raw devices, so that point is moot).

If the /dev/hdc special device file (or /dev/hdc1
or whatever) doesn't exist on your rescue disk, then
(hopefully) "mknod" does exist on your rescue disk,
and you can create the special device files manually

     mknod /hdc  b 22 0
     mknod /hdc1 b 22 1
     mknod /hdc2 b 22 2
     mknod /hdc3 b 22 3
     mknod /hdc4 b 22 4
     mknod /hdc5 b 22 5
     mknod /hdc6 b 22 6
     mknod /hdc7 b 22 7
     mknod /hdc8 b 22 8
     mknod /hdc9 b 22 9

and then run fsck or mount or dd or whatever you
need to do

     e2fsck -fv /hdc1
     mkdir /bar
     mount /hdc1 /bar


Don't fret.  Absolute worst case scenario, you
can simply restore all of your files from last
night's full backup on your off-site triple-
redundant tape drives.


HTH,

D

* On Fri, Oct 27, 2000 at 11:39:31PM -0700, Hawke wrote:
> I've got a bit of a problem folks..
> 
> I ran the hdparm copmmand on my HD and now I cannot get linux
> to boot (it reaches the end of the kernel config and is about to run
> system services when when it states that "error reading from file
> libc.so.6. file data is invalid").
> 
> I cannot seem to get back into my drive to run e2fsck on it,
> the rescue disk I have won't allow me to mount the device
> (in this case /dev/hdc which contains hdc1 hdc2 <hdc5 hdc6>).
> 
> I am at my wits end here and need immediate response (if possible).
> 
> at the moment, I can only use windows 98.
> 
> someone please help!
> 
> Hawke