Date: Tue, 10 Jan 2012 10:30:24 -0800 From: Adrian Chadd <adrian.chadd@gmail.com> To: Joel Dahl <joel@freebsd.org> Cc: freebsd-wireless@freebsd.org Subject: Re: kern/163318: [ath] ath(4) stops working Message-ID: <CAJ-Vmo=XnY5ouzB76Ncpk=UDzRxP8qtm5HNngRs9mXkY9joRnw@mail.gmail.com> In-Reply-To: <201201101800.q0AI0Oxx088371@freefall.freebsd.org> References: <201201101800.q0AI0Oxx088371@freefall.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi, On 10 January 2012 10:00, Joel Dahl <joel@freebsd.org> wrote: > =A0For the record: I put in another PCI ath(4) card and I'm seeing the sa= me > =A0thing with this card. After a while, wireless stops working. It seems = to > =A0go in to some power save mode, and then gets stuck. ifconfig wlan0 sca= n > =A0fixes it, as usual. > > =A0One thing I've noticed though is that it seems to take much longer for= this > =A0card to stop working. Usually around 9-10 hours. Could be a coincidenc= e of > =A0course. Hm. Right, I didn't think it was chipset specific. I think it's very likely another subtle race condition inside net80211 which we just haven't properly fixed. The scan stuff _should_ be running in the net80211 taskqueue (which one exists per interface, iirc.) So it shouldn't be racing with itself, but it may be racing with other bits. I'm trying to reproduce it here, but I won't have a dual-core machine w/ an AR9220 ready for another week or two. Adrian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmo=XnY5ouzB76Ncpk=UDzRxP8qtm5HNngRs9mXkY9joRnw>