> > > hda4: bad access
> > > block 28 count 2 blockend 30 nrsects
> > > end request I/O error dev 03:04 sector 28
> >
I get this too. The "dev 03" part refers to your primary IDE drive.
By running cfdisk under Linux you will probably see something like
the following (note: this is paraphrased from memory):
DEVICE TYPE FS TYPE
/dev/hd0s1 primary Linux
/dev/hd0s2 primary Hurd
/dev/hd0s3 primary Hurd
/dev/hd0s5 Logical Hurd
/dev/hd0s6 Logical Hurd
/dev/hd0s7 Logical Swap
Okay. My setup is a bit weird. I recently replaced a bad disk drive
with a new 4.2 Gig drive. So I partitioned it up into a bunch of < 1Gig
units.
You are limited to Four "Primary" partitions, so the first three are
primary and the last three are "extended" or "logical" drives that
in reality live physically within the final (fourth) un-named primary
partition (/dev/hd0s4).
Under Linux, instead of this error you would see:
Checking IDE drives:
hda1 hda2 hda3 <hda5 hda6 hda7>
I surmise from this that the Hurd (or mach) cannot understand the logical
drives, and so it outputs this error message. E.g., the "04" part of the
error message is the presence of this "wierd" fourth partition that contains
the logical drives.
I believe this is harmless.
-Brent
