I'm not quite sure how attachments work on Launchpad; I couldn't find your attachment for your solution. nonetheless....
I had the same problem, even with 8.04LTS. I can give you my best educated guess. I get this message just after "Will now halt". Looking for that string all over the fs, I found it in /etc/init.d/halt. Right after that, it's attempting to execute halt (as in /sbin/halt). Doing a simple strings(1) on /sbin/halt (actually, it's a symlink to reboot, e.g., /sbin/reboot), it shows /proc/ide in the middle of it. The trouble is, ever since a certain rev. of the kernel, IDE devices are abstracted into SCSI devices; ergo it's no wonder halt cannot iterate over the entries in /proc/ide/., as it will no longer exist. The documentation I've read seems to indicate -h is not really necessary as the kernel will automatically shut down IDE (and network) devices anyway, so it's not necessary (superfluous, redundant) for a program like halt(8) to do this. If anything, halt(8) would have to look at something like /proc/scsi or /proc/bus. My fix/workaround was to edit the halt script so that it NEVER hands "-h" to halt(8). Bingo, no more message. The more troubling one for me is "trying to free already freed IRQ 5". I think what's going on there is similarly halt(8) and the kernel trying to stop all network interfaces. I think it's a bug whereby the 3c509 driver/module tries to shut itself down twice, and doesn't realize it's already shut itself down. (I'm not 100% sure, but I think the card is assigned IRQ 5. too lazy at the moment to confirm :-) .) -- halt：unable to iterate IDE devices:no such files or directory https://bugs.launchpad.net/bugs/193125 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
