On 30.12.2009 18:17, Mark Adler wrote: > The pointer arithmetic underrun reports are correct, but they are > intentional. The reported access overruns are not correct. Mark, thanks for the analysis and your clear explanation! Ralf