wpa_supplicant PSK question

Bob Carlson rjc at sigpro.com
Tue Nov 22 12:07:17 EST 2005


I'm still confused. Using WPA(2)-PSK, the system still needs to carry out
the 4-way handshake. It looks to me like eap_psk.c does that. And that
seems to depend on eap.c and eapol_sm.c. I've probably just missed
something, but I could use a little more help.

Thanks, Bob

> -----Original Message-----
> From: hostap-bounces+bob.carlson=sigpro.com at shmoo.com [mailto:hostap-
> bounces+bob.carlson=sigpro.com at shmoo.com] On Behalf Of Jouni Malinen
> Sent: Monday, November 21, 2005 10:49 PM
> To: hostap at shmoo.com
> Subject: Re: wpa_supplicant PSK question
> 
> WPA-PSK is indeed using EAPOL frames, but it does not require IEEE
> 802.1X/EAPOL or EAP state machines. In other words, you do not need
> eapol_sm.c or eap.c/eap_*.c. One of the easiest ways of finding out what
> is needed is to build wpa_supplicant for Linux and use a minimal
> configuration in .config, e.g.:
> 
> CONFIG_DRIVER_TEST=y
> CONFIG_NO_STDOUT_DEBUG=y
> 
> 
> This links in following files into wpa_supplicant binary:
> 
> config.o eloop.o common.o md5.o rc4.o sha1.o config_file.o base64.o
> l2_packet_linux.o tls_none.o wpa.o preauth.o aes_wrap.o
> wpa_supplicant.o events.o main.o drivers.o driver_test.o
> 
> This is enough to run WPA-PSK, but not WPA-EAP.
> 
> --




More information about the HostAP mailing list