Unfortunately I needed to get the machine back up and so I booted off a livecd, chrooted in and continued the upgrade. I guess it's going to be hard to pin this one down, but I can remember a few points.
At the initramfs shell, there were no hard disk devices at all, and so /dev/disks did not exist. I looked at /proc/modules and saw that the module for my disk controller (via82cxxx) was not loaded. I loaded it with modprobe, and the right kind of kernel messages appeared on the console, but devices for the disks were not created (I don't know if udev should automatically create them in the initramfs environment, or whether I needed to run something else). The basic problem is that the system was left in an unbootable state even though the upgrade was interrupted. I assume this is because something that calls update-initramfs in its postinst was upgraded before the crash, and the resultant initramfs image was unable to find my hard disks. I think the edgy kernel's initramfs should *not* have been touched at all during the upgrade process, or at least a backup should have been made (that is, a single backup at the start of the dist-upgrade, so that if two initramfs-updating packages are upgraded, the second one does not clobber the backup). As for pango, now that the upgrade is complete there is only a single symlink in /etc/pango, named pangox.aliases. It doesn't mention any other files, only x11 font aliases. -- feisty upgrade failure https://bugs.launchpad.net/bugs/108276 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
