krb5_get_server_rcache double free

Sam Hartman hartmans at MIT.EDU
Thu Jul 6 16:55:08 EDT 2006

>>>>> "Rainer" == Rainer Weikusat <rainer.weikusat at> writes:

    Rainer> The krb5_rc_close routine should free rcache, according to
    Rainer> the documentation for this and I of course checked that it
    Rainer> does free rcache. The claim that

    Rainer> 	When krb5_rc_recover_or_initialize() returns failure,
    Rainer> rcache now leaks.

    Rainer> is therefore wrong and the only real effect of this chain
    Rainer> of conversation is restoring the never called free at the
    Rainer> end of the get_server_rcache routine.

    Rainer> <URL:>

Thanks for pointing out the mistake.  We appreciate your contributions
and are going to update the ticket to correct the factual error.

    Rainer> Publishing factually incorrect statements about other
    Rainer> people on the WWW is not so nice.

Agreed.  However we're all human around here and we make mistakes.  I
wish I could say it will not happen again, but it probably will.  No
harm is meant to you.  Even if you had made a mistake and introduced a
memory leak, that wouldn't be a big deal either: humans make those
kind of mistakestoo.

we're happy to append notes to bugs when we make factual errors; you
can also respond and append corrections yourself.

Thanks for your interest in contributing and thanks for pointing out the error.

Sam Hartman
Manager, MIT Kerberos Team

More information about the krbdev mailing list