On Fri, Jun 12, 2020 at 01:08:32PM +0200, Andreas Tobler wrote: > thanks for this patch. We applied it locally and with some modifications we > got it to work. > > The most important modification for us was to move the 'force disconnect > part', the one which follows after the client-addr bind, before we do the > SO_BINDTODEVICE. > If we do a connect with the disconnect_addr which has a sin_family > AF_UNSPEC, we loose the SO_BINDTODEVICE on the socket. > This is the reason we moved it right before we bind the device. > > The second modification is the usage of setsockopt(). Your's was not wrong > but irritating while reading the first time. We just use the devicename and > its length for the 4th and 5th argument instead of the irfeq. > > Attached the updated patch with our modifications. Thanks, applied with fixes and cleanup. -- Jouni Malinen PGP id EFC895FA _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
