The kernel is panicing probably because it isn't finding the correct root. When you switch the drive relationships around it also changes the /dev/hda1 and /dev/hdb3 type stuff.
/dev/hda = primary master
/dev/hdb = primary slave
/dev/hdc = secondary master
/dev/hdd = secondary slave
then partitions for each drive is just a number added on the back of the harddrive name
/dev/hdb1 = first partition on the primary slave IDE device.
You just need to edit the /etc/lilo.conf file or your grub config (you can boot manually from any partition or kernel by using the grub command line at boot up time instead of using the pre-sets you put in it's configuration file) to reflect the change.
Then all you have to do next is manually edit the "/etc/fstab" file.
That is how the OS knows which partition is for what and to know which device is root/swap/cdrom/windows part./zip drive.
Then that will fix it and your OS should then operate 100% back up to snuff.
You may have to update your symbolic links (like shortcuts) to stuff like the /dev/cdrom file which points to your /dev/hd* device that represents the correct cdrom drive.
All in all it pretty simple once you know what to do. The only real problem is getting to the root partition to edit the configuration files and re-run your boot loader...
Most install CDs have a command line or rescue mode that you can access. The CD themselves are usually mini linux distros. Try ctl-alt-F2 to switch to a command line.
IF you can't figure out the install CD, you can create a mini linux distro on a floppy or CDROM like Gentoo/SuSE's live eval CDs or Knoppix CD. Also a usefull thing is
Toms rescue disk
So you boot it up.
Figure out which /dev/hd* is the root partition.
Then make a mount directory and mount it.
mkdir /mnt/rootd
mount /dev/hdb1 /mnt/rootd
Then mount the proc (not always nessicary)
mount -t proc proc /mnt/rootd/proc
CD to the drive and use the chroot command.
cd /mnt/rootd/
chmod /mnt/rootd/
That will get you into your Mandrake Distro. You will still be running the kernel from the rescue disk, but effectively you will be running your "real" install now. You can run commands set up servers and even run X from that chroot enviroment now, but all we want to do is fix the /etc/lilo.conf or grub config, the /etc/fstab file, and rerun the boot loader install command.
Pretty simple stuff after you do it a couple times (or like me pretty often because I screw up a lot of stuff a lot of the time)