On Thu, 2014-06-12 at 14:06 +0200, Thomas Neumann wrote: > Okay. That is definitely odd now. The other output was produced by a 2.6.32 > kernel on SuSE 11SP1. I'm unsure whether to write a small check if $flags > contains rogue 'colon-space-colon-space' sequences or simply ignore the > matter.
I would agree with Thomas Lange that it would be very odd for someone to combine a not-yet-released version of FAI with a lenny or earlier nfsroot. Can we just call it unsupported? > But apart from that: THAT'S AN IDE MACHINE! Great! Can you please provide me > with the output of /proc/ide/*/media? (Another nearly obsolete piece of > cidenow. Even if I force a machine to have IDE disks they're still presented > as SCSI devices.) # ls -l /proc/ide total 0 -r--r--r-- 1 root root 0 2014-06-12 15:27 drivers lrwxrwxrwx 1 root root 8 2014-06-12 15:27 hda -> ide0/hda lrwxrwxrwx 1 root root 8 2014-06-12 15:27 hdc -> ide1/hdc dr-xr-xr-x 3 root root 0 2014-06-12 15:27 ide0 dr-xr-xr-x 3 root root 0 2014-06-12 15:27 ide1 # ls -l /proc/ide/*/media -r--r--r-- 1 root root 0 2014-06-12 15:28 /proc/ide/hda/media -r--r--r-- 1 root root 0 2014-06-12 15:28 /proc/ide/hdc/media # cat /proc/ide/*/media disk cdrom The other machine, which has no CDROM, but does have also SATA disks, has just "disk" for the single IDE disk in /proc/ide/hda . PS I shall send you a bunch of parted outputs via PM soon. Toomas