On 2012-11-10, nuno.silva-uoixanvl...@public.gmane.org wrote: > Hello, > > I am trying to debug a possible issue with my current configuration: I > am using wpa_supplicant together with a broadcom card (through the b43 > driver) under Gentoo. > > The card: > Broadcom Corporation BCM4312 802.11b/g LP-PHY (rev 01) > > And this is under linux version 3.2.1 on amd64, with some Gentoo-level > patches. > > At my school, we have an open wireless network, and computers are > frequently within reach of several access points for this network. > > But the APs seem to employ some load-balancing strategy, where several > APs will turn down the association request claiming they're overloaded > (RX AssocResp with status=17). [...] > Now I don't know if this is some kind of load-balancing or how does it > work, nor do I know if this is the expected behavior (maybe it is?). If > this is not supposed to happen, where can I look to try to fix the > issue? [...]
After updating the kernel (now 3.5.7), I still get similar issues with this driver, in fact, right now I was being connected to a not-so-close AP with bad signal quality, because association with the closest one was rejected (and it still takes some time to try another AP). As I keep both drivers around in my system (just because it never occured to me to uninstall the proprietary since I switched to the open driver), I decided to give the proprietary one a try. So, some rmmods and modprobes later, it seems that the proprietary driver connects *instantly* (that is, to be scientifically accurate, does not take a noticeable amount of time to get connected). That is, for some reason, the open, in-kernel, driver does things in a different way that leads to the scenario I described. So now I guess it would be interesting to somehow compare what both drivers do. As I have access to a wireless network exhibiting the issue and I have both drivers installed, I could easily try to gather some information with each driver. What would be useful for the devs? -- Nuno Silva (aka njsg) http://njsg.sdf-eu.org/ _______________________________________________ b43-dev mailing list b43-dev@lists.infradead.org http://lists.infradead.org/mailman/listinfo/b43-dev
