I am sorry it took me so long to get back to this. I had a good read over the code and this bug and I reached the conclusion that the proposed fix is not 100% adequate.
The issue at stake is a postinst creation of /dev/md* devices by the mdadm package. The original created the devices only when devfs was not in use and /dev/md0 didn't exist already. The current solution creates them if /dev/md0 does not exist. This works, but it's a hack. Namely: - users of devfs will get an error (but who cares?) - udev is somewhat circumvented and MAKEDEV used instead. While this is ugly, it may also come with problems as udev gets more and more important. Thus I propose to change it to use MAKEDEV only if /dev/md0 does not already exist, devfs is not in use, and udev neither, like so: [ ! -e /dev/md0 ] && [ ! -e /dev/.devfsd ] && [ ! -e /dev/.udevdb ] I am going to try this in 2.4.1-1, upgrading and clean installs, with and without udev. If it works for udev, it should work for devfs. Stay tuned. -- Please do not send copies of list mail to me; I read the list! .''`. martin f. krafft <[EMAIL PROTECTED]> : :' : proud Debian developer and author: http://debiansystem.info `. `'` `- Debian - when you have better things to do than fixing a system
signature.asc
Description: Digital signature (GPG/PGP)