> > > /dev/hda1 corresponds to FreeBSDs /dev/ad0s1.
> > > What does Linux call, for example, /dev/ad0s1a?
> > For recent 2.4 kernels, dmesg should have a line of the form:
> > 
> >  p4: <bsd: p11 p12 p13 p14 p15 p16 >
>  hda1: <bsd: hda10 hda11 hda12 hda13 >
>  
> and this is the DragonFly-slice
> 
>  hda2: <bsd: hda14 hda15 hda16 hda17 >
> 
> ..strange numbering, shouldn't it look like this?
>  hda2: <bsd: hda20 hda21 hda22 hda23 >
Nope.  The algorithm Linux uses for reading partition tables etc. is
roughly as follows (from memory):

counter = 0
unexamined_partitions = {entire_disk}
while there are unexamined partitions:
        Pick a partition from the unexamined partitions set
        Assign that partition the number currently in counter
        Look at the partition, and, if it contains any sub-partitions,
                put those in the set of unexamined partitions
        increment counter

What you've seen isn't entirely inconsistent with that, if you have
five logical partitions under the primary for hda4.

> One thing I found particularly interesting is that _both_ /dev/hda2 and
> /dev/hda14 seem to represent /dev/ad0s2a, and I can even mount them both at
> the same time!
This strikes me as an incredibly bad idea.  Definitely don't mount
hda2 read/write. :)

Steven.

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to