2011-05-26 14:29:27

by Arend van Spriel

[permalink] [raw]
Subject: %d in wlan log messages

[ 563.120590] wlan%d: authenticate with c0:c1:c0:04:b8:2c (try 1)
[ 563.128143] wlan%d: authenticated
[ 563.131517] wlan%d: associate with c0:c1:c0:04:b8:2c (try 1)
[ 563.140933] wlan%d: RX AssocResp from c0:c1:c0:04:b8:2c (capab=0x411
status=0 aid=4)
[ 563.148714] wlan%d: associated
[ 563.177019] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready

On latest staging-next I noticed above mentioned log messages. I assume
these message are from cfg80211 or mac80211. Has this been reported?

Gr. AvS

--
Almost nobody dances sober, unless they happen to be insane.
-- H.P. Lovecraft --




2011-05-26 18:01:09

by John W. Linville

[permalink] [raw]
Subject: Re: %d in wlan log messages

On Thu, May 26, 2011 at 07:53:21PM +0200, Arend van Spriel wrote:
> On 05/26/2011 05:09 PM, Daniel Halperin wrote:
> >On Thu, May 26, 2011 at 7:29 AM, Arend van Spriel<[email protected]> wrote:
> >>[ 563.120590] wlan%d: authenticate with c0:c1:c0:04:b8:2c (try 1)
> >>[ 563.128143] wlan%d: authenticated
> >>[ 563.131517] wlan%d: associate with c0:c1:c0:04:b8:2c (try 1)
> >>[ 563.140933] wlan%d: RX AssocResp from c0:c1:c0:04:b8:2c (capab=0x411
> >>status=0 aid=4)
> >>[ 563.148714] wlan%d: associated
> >>[ 563.177019] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
> >>
> >>On latest staging-next I noticed above mentioned log messages. I assume
> >>these message are from cfg80211 or mac80211. Has this been reported?
> >>
> >Interface names [including any numbers] are set by userspace, right?
> >On my Ubuntu 10.04.2 LTS system, it's
> >/etc/udev/rules.d/70-persistent.net.rules (or so, this is from
> >memory). Not a kernel bug.
> >
> >Dan
>
> Weird. I did not change anything in udev rules (using Ubuntu 10.10
> over here). So the only variant in my setup is move to 2.6.39. Can
> you explain that netdev apparently has the proper interface id (see
> timestamp [563.177019] above?

Wasn't a problem like this recently reported w/ compat-wireless?
And the fix was to backport some patch to the networking core?
I could be hallucinating again...

Is it possible that you somehow have a "dirty" build? Could you try
doing a 'make clean' and rebuilding?

John
--
John W. Linville Someday the world will need a hero, and you
[email protected] might be all we have. Be ready.

2011-05-26 15:09:21

by Daniel Halperin

[permalink] [raw]
Subject: Re: %d in wlan log messages

On Thu, May 26, 2011 at 7:29 AM, Arend van Spriel <[email protected]> wrote:
> [ ?563.120590] wlan%d: authenticate with c0:c1:c0:04:b8:2c (try 1)
> [ ?563.128143] wlan%d: authenticated
> [ ?563.131517] wlan%d: associate with c0:c1:c0:04:b8:2c (try 1)
> [ ?563.140933] wlan%d: RX AssocResp from c0:c1:c0:04:b8:2c (capab=0x411
> status=0 aid=4)
> [ ?563.148714] wlan%d: associated
> [ ?563.177019] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
>
> On latest staging-next I noticed above mentioned log messages. I assume
> these message are from cfg80211 or mac80211. Has this been reported?
>

Interface names [including any numbers] are set by userspace, right?
On my Ubuntu 10.04.2 LTS system, it's
/etc/udev/rules.d/70-persistent.net.rules (or so, this is from
memory). Not a kernel bug.

Dan

2011-05-26 18:46:37

by Arend van Spriel

[permalink] [raw]
Subject: Re: %d in wlan log messages

On 05/26/2011 07:55 PM, John W. Linville wrote:
> On Thu, May 26, 2011 at 07:53:21PM +0200, Arend van Spriel wrote:
>>
>> Weird. I did not change anything in udev rules (using Ubuntu 10.10
>> over here). So the only variant in my setup is move to 2.6.39. Can
>> you explain that netdev apparently has the proper interface id (see
>> timestamp [563.177019] above?
> Wasn't a problem like this recently reported w/ compat-wireless?
> And the fix was to backport some patch to the networking core?
> I could be hallucinating again...

Does it happen often then? ;-)

> Is it possible that you somehow have a "dirty" build? Could you try
> doing a 'make clean' and rebuilding?

Ok. I thought I did a distclean before building, but let me try.

Gr. AvS

--
Almost nobody dances sober, unless they happen to be insane.
-- H.P. Lovecraft --



2011-05-26 15:18:49

by Denis 'GNUtoo' Carikli

[permalink] [raw]
Subject: Re: %d in wlan log messages

On Thu, 2011-05-26 at 08:09 -0700, Daniel Halperin wrote:
> are set by userspace, right?
> On my Ubuntu 10.04.2 LTS system, it's
> /etc/udev/rules.d/70-persistent.net.rules (or so, this is from
> memory). Not a kernel bug.
I've the same, with a recent compat wireless(iwconfig returns wlan%d) of
some days ago.

My setup(HTC dream under the SHR GNU/Linux distribution) uses devtmpfs
only tough(and some userspace middleware that also does firmware
loading).

Denis.


2011-05-26 17:53:44

by Arend van Spriel

[permalink] [raw]
Subject: Re: %d in wlan log messages

On 05/26/2011 05:09 PM, Daniel Halperin wrote:
> On Thu, May 26, 2011 at 7:29 AM, Arend van Spriel<[email protected]> wrote:
>> [ 563.120590] wlan%d: authenticate with c0:c1:c0:04:b8:2c (try 1)
>> [ 563.128143] wlan%d: authenticated
>> [ 563.131517] wlan%d: associate with c0:c1:c0:04:b8:2c (try 1)
>> [ 563.140933] wlan%d: RX AssocResp from c0:c1:c0:04:b8:2c (capab=0x411
>> status=0 aid=4)
>> [ 563.148714] wlan%d: associated
>> [ 563.177019] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
>>
>> On latest staging-next I noticed above mentioned log messages. I assume
>> these message are from cfg80211 or mac80211. Has this been reported?
>>
> Interface names [including any numbers] are set by userspace, right?
> On my Ubuntu 10.04.2 LTS system, it's
> /etc/udev/rules.d/70-persistent.net.rules (or so, this is from
> memory). Not a kernel bug.
>
> Dan

Weird. I did not change anything in udev rules (using Ubuntu 10.10 over
here). So the only variant in my setup is move to 2.6.39. Can you
explain that netdev apparently has the proper interface id (see
timestamp [563.177019] above?

Gr. AvS

--
Almost nobody dances sober, unless they happen to be insane.
-- H.P. Lovecraft --



2011-05-26 21:08:38

by Rafał Miłecki

[permalink] [raw]
Subject: Re: %d in wlan log messages

I've heard about this issue once, from someone using brcm80211 of
wl... not sure now.

Could this be brcm80211/wl specific? Just a tip, nothing more than a guess.

--
Rafał