On Sat, Jan 12, 2008 at 08:27:32PM +0000, Da Powah wrote: > > > hi Arno, > > please have a look to my earlier post, because it is marked as spam and you > might not read it. I saw it. > Arno Wagner <wagner@...> writes: > > > Unlikely as this sounds, I think you either changed the passphrase or > > your stored passphrase got corrupted. In both cases there is no > > possible cure than recovering the original passphrase somehow. After > > all, this is the whole point of the encryption. > > > > Arno > > Lets assume my keyfile is ok (i verified it against a backup). > Lets assume, too, the key should unlock the masterkey, because I never played > with luksDelKey / luksAddKey. > > 2 last option may have occured: > #1 a data corruption That would need to be corruption exactly in the key, but nowere else. If it is on disk correct once, it does not net corrupted on disk. Disk ECC is so that undetected errors do not happen in practice. The only way tho get bit-errors is when copying from one disk to another. Did you do that? If not, your LUKS superblock is very likely uncorrupted. > #2 a setup problem, because i played a little and i am always doing the same > nasty failures while trying to uncrypt with Linux LiveCDs. Maybe some Bios > settings harm the setup, too. BIOS should have no influence at all. The liveCD may be the key. So you say when creating a LUKS partition with your normal system and then try to access with a lifeCD, you get the same error? > #1 may it be not, because the mdstat tells me everything is working > an luksDump gives me correct values. > Then #2 is the last possible choice. > Please take a look at my kernel module versions (in my earlier post). Maybe > there are some hints hidden. Sorry, I do not follow the details of the kernel module development. > I am only worrying a bit, because i can encrypt and decrypt a > logical volume on my hda system disk. So the setup seems to be ok > for newly created devices on non-raids. Did you, at any time change the kernel version? If so, mayne get the sources for the version you used originally, compile your own kernel and see whether this works. There may indeed be a bug in some versions of a kernel hash or cipher module. > > thanks for brainstorming with me. No problem. Arno -- Arno Wagner, Dipl. Inform., CISSP --- CSG, ETH Zurich, arno@xxxxxxxxxxx GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F ---- Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans If it's in the news, don't worry about it. The very definition of "news" is "something that hardly ever happens." -- Bruce Schneier --------------------------------------------------------------------- dm-crypt mailing list - http://www.saout.de/misc/dm-crypt/ To unsubscribe, e-mail: dm-crypt-unsubscribe@xxxxxxxx For additional commands, e-mail: dm-crypt-help@xxxxxxxx
