"George R. Ellis" <[EMAIL PROTECTED]> wrote: > The malloc() in eap_tls.c:501 behaves differently on FreeBSD 5.0 than on > Linux (RH 8.0) when (reply->length - TLS_HEADER_LEN) is zero. Under > FreeBSD I end up with a bad address, thus a segmentation fault.
Ok... > This seems to be the problem so far. Does there exist a patch or an > update to this particular problem with malloc() with FreeBSD? Add '+ 1' to the call to malloc. That way, it will always allocate at least one byte of data. Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
