Ok, I now can tell it, it's not the crypto part that has these
problems. After copying the data to an unencrypted disk the
partition couldn't be mounted too. A xfs_repair solved the problem
but I think anyway that the code shouldn't give oops and the mount command
shouldn't end with a seg fault. It's clearly a problem in the XFS code (at
least in kernel 18.104.22.168 & 22.214.171.124).
P.S.: BTW, the xfs_repair even worked at the encrypted partition...
Udo Wolter | /"\
email: uwp@xxxxxxxxxxxxxxxx | \ / ASCII RIBBON CAMPAIGN
www: www.dicke-aersche.de | X AGAINST HTML MAIL
dark: heaven@xxxxxxxxxxxxxx | / \