On 30/01/2008 Marco d'Itri wrote: > > I believe that the bug reported at http://bugs.debian.org/406697 is not > > really an issue with cryptsetup but rather with udev. Can you agree on > > that? If so, I'd like to reassign the bugreport to udev. > > I do not understand what udev should do about this.
Hey Marco, If you read the bugreport log, you might have found the following passage: On Wed, 17 Jan 2007 22:16:17 +0100, David Härdeman wrote: > On Wed, Jan 17, 2007 at 12:15:56PM +0100, Frans Pop wrote: > >I'm now starting to think that the problem could be that the > >'udevd --daemon' process has already been killed during S20sendsigs. > >That would explain that even if events are still generated, they are no > >longer processed and thus the devices remain present. > > > >[...] > > > >I've just verified that udevd is indeed no longer running at the time > >crypto and lvm are shut down. > > Ok, udev creates the /dev/dm-* devices and also removes them. At the > time the cryptdisks scripts are executed at shutdown, udev is not > running (as FJP said). > > The proper "fix" for this seems to be a post-Etch thing (involving > changes in the kernel, udev and libdevmapper), and the issue seems > mostly cosmetical, so I'll mark the bug as such. greetings, jonas
signature.asc
Description: Digital signature
