[krbdev.mit.edu #6277] KLCacheHasValidTickets changed behavior

Alexandra Ellwood via RT rt-comment at krbdev.mit.edu
Tue Dec 16 13:51:40 EST 2008


I'm concerned that this patch does not fix the underlying problem with this function.   The patch 
changes the default return value from valid to expired.  However the code below is *supposed* 
to catch all non-valid ticket conditions.  I'm not disputing that there is a real bug here, but 
changing the default return value is probably just masking the real bug where one of the non-
valid ticket conditions is not being detected by the logic.

In particular using this method of fixing the bug may mean that KIM returns that tickets are 
expired when they are invalid for some other reason (need validation/not yet valid).

Would it be possible to get a description of the state of the ccache when KLCacheHasValidTickets returned that tickets were valid even though they weren't?   I looked for 
a regression test flagged with this bug number and didn't see one.



More information about the krb5-bugs mailing list