Date: Sun, 15 Jul 2007 21:15:58 -0700 From: "Bruce A. Mah" <bmah@freebsd.org> To: Sam Leffler <sam@errno.com> Cc: freebsd-mobile@freebsd.org Subject: Re: ath(4), wpa_supplicant, WPA2, Netgear WG302 problem Message-ID: <469AF0FE.1060400@freebsd.org> In-Reply-To: <46994990.1070505@errno.com> References: <20070713202015.GA1718@phantom.kitchenlab.org> <46994990.1070505@errno.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --] If memory serves me right, Sam Leffler wrote: > Bruce A. Mah wrote: >> The OS is FreeBSD HEAD as of yesterday, GENERIC kernel. Note that >> this has the recent HAL import, as well as wpa_supplicant v0.5.8: >> >> ath_hal: 0.9.20.3 (AR5210, AR5211, AR5212, RF5111, RF5112, RF2413, RF5413) > > not sure what "the recent hal import" is My bad. I somehow thought there was a HAL import associated with the background scanning update you did recently. [snip] > It appears wpa_supplicant is timing out waiting for an initial EAPOL > frame from the ap. I'd verify traffic is getting through; possibly with > tcpdump. This means I should be running tcpdump on a different machine with a wireless interface in monitor mode, right? The AP's log file shows that the card is authenticating and deauthenticating immediately (so the AP at least knows of the card's existence). I haven't compared this with the logs of a successful assocation yet. Also I tried this experiment with a WG511T from my wife's laptop and got the same results. > BTW I'd expect more verbose debug output from wpa_supplicant > with -dd. That was just the first iteration of attempting to associate, but subsequent iterations (except for finding different APs in the AP scan) looked identical to me. I've gotten the card to associate a few times, but generally after many tens of minutes. Once I get an association, the card can pass traffic without any particular problem. (And I did get a *lot* more verbose output from wpa_supplicant the one time that I had invoked it manually with -dd, unfortunately I wasn't saving this at the time.) I've also noticed occasional "interrupt storm" messages for the irq where the cardbus bridge is. I tried upping the limit from 1000 to 5000, but this didn't always get rid of the errors. I'm not sure if this is significant or not. This irq was shared with the firewire controller on a GENERIC kernel; I removed the firewire driver from the kernel hoping this would help and ended up with fxp0 sharing the irq. Not sure if this is a productive line of investigation. I'll continue to play with this in my Copious Spare Time (TM)... Thanks for your advice! Bruce. [-- Attachment #2 --] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGmvD+2MoxcVugUsMRAgABAKCCQeKDvWTpFgH07W3gV8DKiMG4iQCeM9vj pB1hIZd7vClK3OsnUp8CCKg= =LIb8 -----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?469AF0FE.1060400>