As I am going on a holiday, I am currently ripping all my CDs using grip, and last night I experienced a race condition, probably in the kernel or supermount. What I did was simply [root@home mp3]# df Filesystem 1k-blocks Used Available Use% Mounted on /dev/hda5 248932 83908 165024 34% / none 192212 0 192212 0% /dev/shm /dev/hda10 8385636 7173588 1212048 86% /home and here the process locked up (next device in list should be /mnt/cdrom), and it is impossible to kill it (kill -9 is no good), and I can't use cdrom drive again before rebooting :( The reason it is locked up is most likely because I was ejecting the CD at approx. the same time, so it might have got an OK in the beginning of hte request, and then times out later. I think this is a bad one, and a symptom of some race condition in ther kernel / supermount.
