I'd like to use the Atheros chipset in a master mode that performs =20 802.1x authentication. Although hostapd is currently targeted only at =20 the hostap driver, I think the problem of making it more generic is not =20 insurmountable. Are there other folks interested/actively involved in =20 this effort? I don't want to duplicate effort or end up with something =20 that won't ever get picked up if I venture into doing. As far as I can tell, the 802.1x subsystem of hostapd depends on the =20 hostap driver for only a few non-standard features. There's a lot of =20 room in hostapd to change the API to driver.c into something that's =20 more hardware neutral. Apart from a couple wierd calls the WPA code =20 uses, there are calls for enabling "HOSTAPD mode" and tuning 802.1x =20 mode on, on and off per-station encryption, and toggling host-based =20 encryption. It makes sense (to me) to implement versions of the following =20 functionality for madwifi: - Per-client "authorized" flag, instead of (what appears to be) M_HOSTAP authorizing all clients blindly. This allows a fast path through the driver for forwarding frames from associated hosts. - Per-client keys. This is implemented in HostAP by storing per-associated-client state in the driver containing each of the tx/rx keys. Again, this is useful in order to provide fast-path bridging. There are get/set commands through the PRISM API for per- host keys to be pushed/pulled from the driver There are individaul REMOVE_STA and FLUSH command to selectively or completely clean up this state. - The above implies encryption is done in software, that but that seems to be the current case with madwifi, modulo the current lack in flexibility in which keys to encrypt with. - Paths for sending raw frames to hostapd. Specifically, PAE, management, and "To DS" frames from non-associated/unknown stations. The code in ath_rx_capture() is pretty close to what's needed. - Apart from a couple wierd calls the WPA code uses, there are calls for enabling "HOSTAPD mode" and tuning 802.1x mode on (mainly providing the functionality above) turning on and off per-station encryption, and toggling host-based encryption. I'm fishing for reactions.. -- Paul ------------------------------------------------------- This SF.Net email is sponsored by: SourceForge.net Broadband Sign-up now for SourceForge Broadband and get the fastest 6.0/768 connection for only $19.95/mo for the first 3 months! http://ads.osdn.com/?ad_id=2562&alloc_id=6184&op=click _______________________________________________ Madwifi-devel mailing list Madwifi-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/madwifi-devel
| Andrew Morton | Re: Linux 2.6.21-rc4 |
| Andrew Morton | -mm merge plans for 2.6.23 |
| Greg KH | [GIT PATCH] driver core patches against 2.6.24 |
| Balbir Singh | Re: [RFC][PATCH 2/7] RSS controller core |
git: | |
| Gerrit Renker | [PATCH 15/37] dccp: Set per-connection CCIDs via socket options |
| David Miller | [GIT]: Networking |
| Andreas Henriksson | [PATCH 06/12] Remove bogus reference to tc-filters(8) from tc(8) manpage. |
| Jarek Poplawski | Re: [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
