lib/ldb/tests: add test for ldb.Dn passed utf8 unicode

lib/ldb/tests: add test for ldb.Dn passed utf8 unicode

object dn format should be a utf8 encoded string
Note: Currently this fails in python2 as the c python binding for
      the dn string param uses PyArg_ParseTupleAndKeywords() with 's'
      format, this will accept str *or* unicode in the default encoding.
      The default encoding in python2 is... ascii.

Also adding here a knownfail to squash the error produced by the test.

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>