Re: IDE disk geometry + patch

Nahshon (
Sun, 6 Dec 1998 20:47:34 +0200 (EET)

I got a new large IDE disk (>8GB) and was worried about how
it would be handled by Linux/fdisk. (before this thread
began) I have set it to LBA. Fdisk (tried on Linux 2.0.35
and 2.1.1XX) reports C/H/S = 2100/255/63 (which matches
my calculation). I guess these are the values returned by
the BIOS. I did not see any problems when using LBA and
fdisk (am I just lucky).

On another instance, A friend has partitioned his large disk
using a commercial software. We ran Linux fdisk (from RH 5.1)
just to verify the partitions and to set partition types.
Fdisk reported the right number of cylinders for the "virgin"
disk but only 1023 cylinders after the disk was partitioned.
That makes me question where the disk geometry is realy taken
from (is it read of the disk or reported by the firmware?).

Are the problems described with LBA generic, or a result of
broken/incompatible bioses?
I can easily repartition my disk (nothing important on it yet).
Should I switch back from LBA? What can I gain except for
maybe 0.1% of the disk space?
What are the current limits for Lilo or Grub about the location
of boot images? I keep a /boot primary partition on the first
LBA cylinder and use lilo or grub. Linux is installed on logical
partitions which go beyond the 8GB mark. I plan to install
other OS.

Thank you very much,
Itai Nahshon


- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to Please read the FAQ at