Problems with SATA and >15 partitions
Hi,
recently we (debian-amd64) have gotten some problem reports for SATA.
Depending on the kernel version, how it is configured and how the
initrd is build an SATA disk shows up as /dev/hde or /dev/sda.
This has three consequences:
1. going back and forth between an old and new kernel means changing
the fstab (for people without LABEL / UUID).
2. The D-I can mess up and build an initrd for /dev/hde that actually
has /dev/sda (or vice versa, can't remember).
3. People with more than 15 partition on /dev/hde are unable to reach
all partition on /dev/sda.
I don't think there is much we can do about 1 except decide which
driver should be prefered and make sure that Debian kernels will stick
with it (I would suggest the scsi SATA driver and /dev/sda since thats
the better one).
For 2 this might be a bug in mkinitrd. I don't have an SATA disk so it
would be good if another developer with such a disk could look into
this and try the various upgrade paths: kernel with ide SATA active to
kernel-image deb with scsi SATA and vice versa. Upgrade from 2.4 to
2.6 and back. Maybe mkinitrd should detect when the device has to
change and warn the user.
As for 3 this is a general problem of scsi disks. Wouldn't it be
possible for the code that scans partitions to allocate device nodes
dynamically for partitions >15 so that users with devfs or udev active
would get additional device nodes created? Alternatively a userspace
program could scan the partition block and create devices with the
device mapper coresponding to each partition. Both aproaches would
allow for >15 partition.
Ideas, comments, experience?
MfG
Goswin
Reply to: