[ath9k-devel] ar9271 SIOCSIFFLAGS error and authentication timeout

[Cedric Sodhi <manday@gmx.net>]

I've just inserted a makeshift "rfkill unblock 0" in the init script
which got me rid of the two SIOCSIFFLAGS errors. But I still get an
authentication timed out, and apart from that nothing changed. I guess
there has to be something deeper than just a flawed init script.

I can't rule out that I have made a kernel misconfig apart from the
explicit settings which are stated to be required for compat-wireless.


On Thu, Feb 24, 2011 at 05:21:34PM +0100, Cedric Sodhi wrote:
> I'm not sure whether I understand every bit of your reply. For practical
> reasons, I would like to skip directly to...
> 
> On Thu, Feb 24, 2011 at 04:08:15PM +0200, Jouni Malinen wrote:
> > On Thu, 2011-02-24 at 03:52 -0800, Peter Stuge wrote:
> > > Quick grep rfkill -r /lib/rc* /etc/init.d comes up empty.
> > 
> > It is done with iwconfig. iwconfig_post_stop and failure in
> > iwconfig_pre_start results in "iwconfig wlan# txpower off" getting
> > executed. This results in soft blocking the interface.
> > 
> > iwconfig_pre_starts removes the soft block with iwconfig_defaults that
> > runs "iwconfig wlan# txpower auto", but this happens only after
> > wpa_supplicant has been started (I would do that in different order..)
> > and as such, depends on wpa_supplicant being able to handle a dynamic
> > rfkill blocking change.
> > 
> > Anyway, this is getting way off-topic for this mailing list. Like I
> > mentioned before, if this soft blocking causes issues, the question
> > should be directed to Gentoo maintainers. Though, it should be noted
> > that this works fine on my laptop with Gentoo as long as the network
> > settings are correct to avoid hitting the soft block on failure and as
> > long as wpa_supplicant is recent enough to handle the rfkill events
> > properly.
> 
> So you are implying that it's actually an issue of my network settings
> and/or wpa_supplicant version? Is this the quintessence?
> 
> If so, what networksettings do I require and what version of
> wpa_supplicant do you use?
> 
> Thanks, it sounded like it's already solved
> 
> > In addition, it should be noted that any manual operation
> > after "net.wlan# stop" will need to take care of unblocking the soft
> > block before actually trying to use the radio.
> > 
> > - Jouni
> 
> There, I lost you again. After the stop, there is no soft lock in place,
> why are you saying this then?
> 
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel