kerberos, pre_auth, and smartcards

[Russ Allbery]

Greg Hudson <ghudson at MIT.EDU> writes:

> preauth-required doesn't specify which kind of preauth is required.  The
> client is proving its knowledge of the password using a much simpler
> mechanism called "encrypted timestamp", and that's sufficient for the
> KDC to issue a ticket.

> Currently, if you want to specifically require pkinit, you'll need to
> randomize the principal's key so that there is no valid password.

I thought setting requires_hwauth on the principal should force PKINIT.
Does this not work the way that I thought it did?

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>