[pkg-wpa-devel] Bug#542917: iwl3945: iwl3945 fickly connect

Ben Hutchings ben at decadent.org.uk
Wed Sep 2 12:16:04 UTC 2009


On Wed, 2009-09-02 at 19:09 +1000, Kel Modderman wrote:
> On Wednesday 02 September 2009 10:09:20 Ben Hutchings wrote:
> > On Tue, 2009-09-01 at 19:36 +0200, Daniel Huhardeaux wrote:
> > > Ben Hutchings a écrit :
> > > > On Sat, 2009-08-22 at 11:07 +0200, Daniel wrote:
> > > >> Package: linux-image-2.6.30-1-amd64
> > > >> Version: 2.6.30-6
> > > >> Severity: important
> > > >> File: iwl3945
> > > >>
> > > >>
> > > >> 0b:00.0 Network controller: Intel Corporation PRO/Wireless 3945ABG [Golan] Network Connection (rev 02)
> > > >>
> > > >> Having the above wireless device, I always have problems with connection to AP using wpa_supplicant.
> > > > 
> > > > When you say "always", do you mean you have seen this problem with
> > > > earlier kernel versions as well?
> > > 
> > > Yes. See
> > > 
> > > http://lists.debian.org/debian-kernel/2009/03/msg00123.html
> > > 
> > > and Bug #12880 on bugzilla.kernel.org
> > 
> > That's a different bug, reported as <http://bugs.debian.org/518358>,
> > which you said was fixed in version 2.6.29-2.
> > 
> > I'm talking about the bug reported as <http://bugs.debian.org/542917>.
> > 
> > > > How are you invoking wpa_supplicant?
> > > Started from /etc/netwok/interfaces and/or wpa_gui
> > > 
> > >    Do you use a front-end such as
> > > > network-manager or wicd?
> > > No
> > 
> > I've looked at where the message "wlan0: deauthenticating by local
> > choice (reason=3)" is generated, and it appears to mean that something
> > is setting the interface down.  I have been using the same WLAN device
> > and the same kernel version as you with no such problems.  So I think
> > this is an issue with wpa_supplicant or wpa_gui, or the way you have
> > configured them.  I will reassign the bug accordingly.
> 
> I'd say the deauth message is appearing and _then_ the interface is taken down
> in reaction.

The deauth message comes from:

        int ieee80211_sta_deauthenticate(struct ieee80211_sub_if_data *sdata, u16 reason)
        {
        	printk(KERN_DEBUG "%s: deauthenticating by local choice (reason=%d)\n",
        	       sdata->dev->name, reason);
        
which is called from the close/down function for mac80211 interfaces:
        
        static int ieee80211_stop(struct net_device *dev)
        {
        ...
        	if (sdata->vif.type == NL80211_IFTYPE_STATION) {
        		if (sdata->u.mgd.state != IEEE80211_STA_MLME_DISABLED)
        			ieee80211_sta_deauthenticate(sdata,
        				WLAN_REASON_DEAUTH_LEAVING);
        	}

WLAN_REASON_DEAUTH_LEAVING is defined as 3.

It can also be called in response to a command sent over netlink.

> wpa_supplicant simply reacts to events the kernel driver gives it.
> wpa_supplicant doesn't look suspicious here at all given this information.

What about wpa_gui?

Ben.

-- 
Ben Hutchings
If at first you don't succeed, you're doing about average.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.alioth.debian.org/pipermail/pkg-wpa-devel/attachments/20090902/99ae6d6a/attachment-0002.pgp>


More information about the Pkg-wpa-devel mailing list