SEGV in krb5_free_cred_contents

Markus Moeller huaraz at moeller.plus.com
Sat Oct 24 17:27:31 EDT 2009


"Ezra Peisach" <epeisach at MIT.EDU> wrote in message 
news:4AE2F442.80202 at mit.edu...
>
> a) Is this a multithreaded application?
> b) Are you sharing the same cache between threads?  If so - destroying
> one while in use can cause this.

It is not threaded.

> c) Can you provide some sample code - if not - can you indicate how you
> are resolving the memory cache? Are you specifying "MEMORY:" or
> "MEMORY:fooo" (there is an app out there that assumes that "MEMORY:"

I used MEMORY:mem_cache

I will see if I can get a small sample app.

> will give a unique cache - which it doesn't - so it dies in a
> multithreaded app).
> d) If you have access to a linux box - valgrind is very useful for
> tracking down these problems...
> e) I suppose there could be a difference in OpenSolaris vs MIT code -
> you can trying to compile MIT's code base and see if the problem exists...
>

It looks like a OpenSolaris problem as I don't see it on Linux.

> Ezra
>

Thank you
Markus
> ________________________________________________
> Kerberos mailing list           Kerberos at mit.edu
> https://mailman.mit.edu/mailman/listinfo/kerberos
> 





More information about the Kerberos mailing list