Re: (un)corrupted ext2 partitions

Guest section DW (dwguest@win.tue.nl)
Thu, 14 Jan 1999 13:47:11 +0100 (MET)


From: Arvind Sankar <arvinds@mit.edu>

> From: Arvind Sankar <arvinds@MIT.EDU>
>
> Does the ext2 file system depend on /dev/hda5 staying /dev/hda5?
>
> No; but it does depend on the starting location of /dev/hda5 staying
> constant if you play partition renaming games. This isn't always the
> case with extended partitions, though, because of their really gross
> nesty partition table structure, so you have to exercise a lot of care
> when you move extended partitions around. I generally will try very
> hard to avoid extended partitions whenever posible.
>
> - Ted

Interesting, that about extended partitions. Anyway, I need them, cos I
got DOS, linux and netbsd on my disk, and I can't squeeze all of them into
the four primary partitions.
Are you saying that though fdisk says /dev/hda5 starts at the same cylinder
as my original /dev/hda6 started, it might actually be different at the
sector level? (my disk is LBA, of course)

I missed the start of this discussion and don't know
what you are talking about. But yes, logical partitions
do not usually start at a cylinder boundary, so if you
mess around with partitions you have to make sure that
everything is OK on the sector level.

You can see sector details using the "cfdisk -Ps /dev/hda" command.
It will tell you the offset in the extended partition where the
logical partition starts.

On the other hand, if I am not mistaken Ted was talking about
something else, namely the fact that logical partitions do not
have an absolute address, but a relative address w.r.t. to previous
logical partition. Thus, if you delete the previous one by hand
you are asking for trouble.

The command "cfdisk -Pt /dev/hda" will show such details.

Andries

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/