[krbdev.mit.edu #6787] S4U memory leak

Luke Howard via RT rt-comment at krbdev.mit.edu
Sat Oct 9 09:47:05 EDT 2010


[ghudson - Fri Oct  8 17:43:18 2010]:

> > This looks OK, but instead of adding a bitfield to indicate a memory
> > ccache, use krb5_cc_get_type().
> 
> Wouldn't that do the wrong thing if the cred ccache is a pre-existing
> memory ccache which the caller still has interest in?

Hmm. Fair enough.



More information about the krb5-bugs mailing list