On Tue, Jan 07, 2020 at 11:37:05AM +0100, Alexander Wetzel wrote: > Am 07.01.20 um 01:01 schrieb "Jürgen Bausa": > > The log can be found at: > > > > https://pastebin.com/CLA7gx4g > > > > The computer resumes at 00:35:46 but fails to connect. At 00:36:18 the second attempt to > > connect is successfull. > As expected we get the "official" completion messages: > Jan 6 00:35:51 lina wpa_supplicant[662]: wlp1s0: CTRL-EVENT-CONNECTED - > Connection to 98:9b:cb:12:c0:1c completed [id=0 id_str=] At this point, everything was completed successfully as far as wpa_supplicant is concerned.. > But then the AP starts the eapol handshake again, without any obvious reason > for us (It's allowed to do so, but unusual): > > Jan 6 00:35:51 lina wpa_supplicant[662]: l2_packet_receive: > src=98:9b:cb:12:c0:1c len=99 > Jan 6 00:35:51 lina wpa_supplicant[662]: wlp1s0: RX EAPOL from > 98:9b:cb:12:c0:1c The most common case for seeing another EAPOL-Key message here is a retransmission of EAPOL-Key msg 3/4, but this one is actually msg 1/4. That looks pretty strange. > Continuing here and figure out what exactly is wrong will be harder. > Ideally we could get a log if similar detail level from your router but > chances are it will not add much to the picture. A independent sniffer > capturing the handshake (attempts) would be able to confirm which end of the > communication is the failing one, but that's a complex topic if you never > have done that. I don't see much of an other options for figuring out what is failing here than getting that sniffer capture and ideally a debug log from the AP. Since everything looks fine from wpa_supplicant view point, there might be something else to be gained from the driver debug log, but I'm not optimistic on that explaining much more without the information from external sources since decision to send out msg 1/4 during the same association is fully within the AP and there was no indication of any trigger that the station could have done to get that second handshake started. -- Jouni Malinen PGP id EFC895FA _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
