[ath9k-devel] Can't associate with a particular AP

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

[ please CC me when replying, I am not subscribed to the list ]

Hi,

I have recently purchased a new laptop (Asus X93S) with an Atheros AR9285
wireless chip which obviously uses the ath9k module.

When connecting to the AP of my router, everything works as expected
provided no other PC has already been connected to the AP. If that is not
the case, I have to reboot the router and make sure my new laptop is the
first to connect.

Here is the log of such a failing connection:
    Oct  1 08:55:41 ares kernel: [ 8774.854934] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
    Oct  1 08:55:41 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
    Oct  1 08:55:48 ares kernel: [ 8781.880987] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
    Oct  1 08:55:49 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
    Oct  1 08:55:49 ares kernel: [ 8781.931122] cfg80211: Calling CRDA to update world regulatory domain
    Oct  1 08:55:49 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
    Oct  1 08:55:49 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
    Oct  1 08:55:50 ares wpa_supplicant[1859]: Trying to authenticate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
    Oct  1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: scanning -> authenticating
    Oct  1 08:55:50 ares kernel: [ 8783.091580] wlan0: authenticate with ce:67:c3:86:34:80 (try 1)
    Oct  1 08:55:50 ares wpa_supplicant[1859]: Trying to associate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
    Oct  1 08:55:50 ares kernel: [ 8783.093674] wlan0: authenticated
    Oct  1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: authenticating -> associating
    Oct  1 08:55:50 ares kernel: [ 8783.114042] wlan0: associate with ce:67:c3:86:34:80 (try 1)
    Oct  1 08:55:50 ares kernel: [ 8783.123261] wlan0: RX ReassocResp from ce:67:c3:86:34:80 (capab=0x411 status=0 aid=2)
    Oct  1 08:55:50 ares kernel: [ 8783.123265] wlan0: associated
    Oct  1 08:55:50 ares wpa_supplicant[1859]: Associated with ce:67:c3:86:34:80
    Oct  1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
    Oct  1 08:55:52 ares kernel: [ 8785.016027] wlan0: no IPv6 routers present
    Oct  1 08:55:56 ares kernel: [ 8789.887920] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
    Oct  1 08:55:56 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
    Oct  1 08:55:56 ares kernel: [ 8789.903098] cfg80211: Calling CRDA to update world regulatory domain
    Oct  1 08:55:56 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
    Oct  1 08:55:57 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
    Oct  1 08:55:58 ares wpa_supplicant[1859]: Trying to authenticate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
    Oct  1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: scanning -> authenticating
    Oct  1 08:55:58 ares kernel: [ 8791.063535] wlan0: authenticate with ce:67:c3:86:34:80 (try 1)
    Oct  1 08:55:58 ares wpa_supplicant[1859]: Trying to associate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
    Oct  1 08:55:58 ares kernel: [ 8791.070815] wlan0: authenticated
    Oct  1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: authenticating -> associating
    Oct  1 08:55:58 ares kernel: [ 8791.091384] wlan0: associate with ce:67:c3:86:34:80 (try 1)
    Oct  1 08:55:58 ares kernel: [ 8791.100579] wlan0: RX ReassocResp from ce:67:c3:86:34:80 (capab=0x411 status=0 aid=2)
    Oct  1 08:55:58 ares kernel: [ 8791.100583] wlan0: associated
    Oct  1 08:55:58 ares wpa_supplicant[1859]: Associated with ce:67:c3:86:34:80
    Oct  1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
    Oct  1 08:56:04 ares kernel: [ 8797.894857] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
    Oct  1 08:56:05 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
    Oct  1 08:56:05 ares kernel: [ 8797.933108] cfg80211: Calling CRDA to update world regulatory domain
    Oct  1 08:56:05 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
    Oct  1 08:56:05 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
    Oct  1 08:56:06 ares NetworkManager[1830]: <warn> Activation (wlan0/wireless): association took too long.
    Oct  1 08:56:06 ares NetworkManager[1830]: <info> (wlan0): device state change: config -> need-auth (reason 'none') [50 60 0]
    Oct  1 08:56:06 ares NetworkManager[1830]: <warn> Activation (wlan0/wireless): asking for new secrets

lspci output:
    03:00.0 Network controller: Atheros Communications Inc. AR9285 Wireless Network Adapter (PCI-Express) (rev 01)
        Subsystem: AzureWave Device 1089
        Flags: bus master, fast devsel, latency 0, IRQ 17
        Memory at f7900000 (64-bit, non-prefetchable) [size=64K]
        Capabilities: [40] Power Management version 3
        Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit-
        Capabilities: [60] Express Legacy Endpoint, MSI 00
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [160] Device Serial Number 00-15-17-ff-ff-24-14-12
        Capabilities: [170] Power Budgeting <?>
        Kernel driver in use: ath9k

iwconfig:
    wlan0     IEEE 802.11bgn  ESSID:"Kirya"
              Mode:Managed  Frequency:2.422 GHz  Access Point: C6:6E:27:A6:BE:A4
              Bit Rate=65 Mb/s   Tx-Power=14 dBm
              Retry  long limit:7   RTS thr:off   Fragment thr:off
              Encryption key:off
              Power Management:off
              Link Quality=70/70  Signal level=-29 dBm
              Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
              Tx excessive retries:2  Invalid misc:90   Missed beacon:0

All my laptops run Debian unstable, currently with kernel 3.1~rc7 (from
Debian).

I have already tested changing the security scheme from WPA (AES) to WPA
(AES/TKIP), WPA (TKIP), WEP and even without security, but the problem is
the same.

I own other laptops using the brcmsmac and b43 modules which do not have any
problem with this router. On the other hand, I have successfully tested the
laptop chip with other routers with success.

I normally use NetworkManager as a wireless client, but have also tested
wicd and manual configuration (through Debian ifupdown and wpa_supplicant),
but still encounter the same problem.

I do not have much control on the router as it is the one supplied by my
internet provider, hence I cannot access to the logs.

I suspect it might be a problem of the ath9k module and/or in the router
itself, but I do not know how to confirm any of these speculations.

I have already noticed similar issues [0] in the archive of the list, which
lets me think I am in the right direction, but what extra information would
you need to help you understand the issue?

Cheers,
Julien

[0] https://lists.ath9k.org/pipermail/ath9k-devel/2011-October/007055.html

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111003/c4772ddb/attachment-0001.pgp 

[ath9k-devel] Can't associate with a particular AP

[Mohammed Shafi]

On Mon, Oct 3, 2011 at 10:06 PM, Julien Valroff <julien@kirya.net> wrote:
> [ please CC me when replying, I am not subscribed to the list ]
>
> Hi,
>
> I have recently purchased a new laptop (Asus X93S) with an Atheros AR9285
> wireless chip which obviously uses the ath9k module.
>
> When connecting to the AP of my router, everything works as expected
> provided no other PC has already been connected to the AP. If that is not
> the case, I have to reboot the router and make sure my new laptop is the
> first to connect.
>
> Here is the log of such a failing connection:
> ? ?Oct ?1 08:55:41 ares kernel: [ 8774.854934] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
> ? ?Oct ?1 08:55:41 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
> ? ?Oct ?1 08:55:48 ares kernel: [ 8781.880987] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
> ? ?Oct ?1 08:55:49 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
> ? ?Oct ?1 08:55:49 ares kernel: [ 8781.931122] cfg80211: Calling CRDA to update world regulatory domain
> ? ?Oct ?1 08:55:49 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
> ? ?Oct ?1 08:55:49 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
> ? ?Oct ?1 08:55:50 ares wpa_supplicant[1859]: Trying to authenticate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
> ? ?Oct ?1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: scanning -> authenticating
> ? ?Oct ?1 08:55:50 ares kernel: [ 8783.091580] wlan0: authenticate with ce:67:c3:86:34:80 (try 1)
> ? ?Oct ?1 08:55:50 ares wpa_supplicant[1859]: Trying to associate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
> ? ?Oct ?1 08:55:50 ares kernel: [ 8783.093674] wlan0: authenticated
> ? ?Oct ?1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: authenticating -> associating
> ? ?Oct ?1 08:55:50 ares kernel: [ 8783.114042] wlan0: associate with ce:67:c3:86:34:80 (try 1)
> ? ?Oct ?1 08:55:50 ares kernel: [ 8783.123261] wlan0: RX ReassocResp from ce:67:c3:86:34:80 (capab=0x411 status=0 aid=2)
> ? ?Oct ?1 08:55:50 ares kernel: [ 8783.123265] wlan0: associated
> ? ?Oct ?1 08:55:50 ares wpa_supplicant[1859]: Associated with ce:67:c3:86:34:80
> ? ?Oct ?1 08:55:50 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
> ? ?Oct ?1 08:55:52 ares kernel: [ 8785.016027] wlan0: no IPv6 routers present
> ? ?Oct ?1 08:55:56 ares kernel: [ 8789.887920] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
> ? ?Oct ?1 08:55:56 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
> ? ?Oct ?1 08:55:56 ares kernel: [ 8789.903098] cfg80211: Calling CRDA to update world regulatory domain
> ? ?Oct ?1 08:55:56 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
> ? ?Oct ?1 08:55:57 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
> ? ?Oct ?1 08:55:58 ares wpa_supplicant[1859]: Trying to authenticate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
> ? ?Oct ?1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: scanning -> authenticating
> ? ?Oct ?1 08:55:58 ares kernel: [ 8791.063535] wlan0: authenticate with ce:67:c3:86:34:80 (try 1)
> ? ?Oct ?1 08:55:58 ares wpa_supplicant[1859]: Trying to associate with ce:67:c3:86:34:80 (SSID='Kirya' freq=2422 MHz)
> ? ?Oct ?1 08:55:58 ares kernel: [ 8791.070815] wlan0: authenticated
> ? ?Oct ?1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: authenticating -> associating
> ? ?Oct ?1 08:55:58 ares kernel: [ 8791.091384] wlan0: associate with ce:67:c3:86:34:80 (try 1)
> ? ?Oct ?1 08:55:58 ares kernel: [ 8791.100579] wlan0: RX ReassocResp from ce:67:c3:86:34:80 (capab=0x411 status=0 aid=2)
> ? ?Oct ?1 08:55:58 ares kernel: [ 8791.100583] wlan0: associated
> ? ?Oct ?1 08:55:58 ares wpa_supplicant[1859]: Associated with ce:67:c3:86:34:80
> ? ?Oct ?1 08:55:58 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associating -> associated
> ? ?Oct ?1 08:56:04 ares kernel: [ 8797.894857] wlan0: deauthenticated from ce:67:c3:86:34:80 (Reason: 15)
> ? ?Oct ?1 08:56:05 ares wpa_supplicant[1859]: CTRL-EVENT-DISCONNECTED bssid=ce:67:c3:86:34:80 reason=15
> ? ?Oct ?1 08:56:05 ares kernel: [ 8797.933108] cfg80211: Calling CRDA to update world regulatory domain
> ? ?Oct ?1 08:56:05 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: associated -> disconnected
> ? ?Oct ?1 08:56:05 ares NetworkManager[1830]: <info> (wlan0): supplicant interface state: disconnected -> scanning
> ? ?Oct ?1 08:56:06 ares NetworkManager[1830]: <warn> Activation (wlan0/wireless): association took too long.
> ? ?Oct ?1 08:56:06 ares NetworkManager[1830]: <info> (wlan0): device state change: config -> need-auth (reason 'none') [50 60 0]
> ? ?Oct ?1 08:56:06 ares NetworkManager[1830]: <warn> Activation (wlan0/wireless): asking for new secrets
>
> lspci output:
> ? ?03:00.0 Network controller: Atheros Communications Inc. AR9285 Wireless Network Adapter (PCI-Express) (rev 01)
> ? ? ? ?Subsystem: AzureWave Device 1089
> ? ? ? ?Flags: bus master, fast devsel, latency 0, IRQ 17
> ? ? ? ?Memory at f7900000 (64-bit, non-prefetchable) [size=64K]
> ? ? ? ?Capabilities: [40] Power Management version 3
> ? ? ? ?Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit-
> ? ? ? ?Capabilities: [60] Express Legacy Endpoint, MSI 00
> ? ? ? ?Capabilities: [100] Advanced Error Reporting
> ? ? ? ?Capabilities: [140] Virtual Channel
> ? ? ? ?Capabilities: [160] Device Serial Number 00-15-17-ff-ff-24-14-12
> ? ? ? ?Capabilities: [170] Power Budgeting <?>
> ? ? ? ?Kernel driver in use: ath9k
>
> iwconfig:
> ? ?wlan0 ? ? IEEE 802.11bgn ?ESSID:"Kirya"
> ? ? ? ? ? ? ?Mode:Managed ?Frequency:2.422 GHz ?Access Point: C6:6E:27:A6:BE:A4
> ? ? ? ? ? ? ?Bit Rate=65 Mb/s ? Tx-Power=14 dBm
> ? ? ? ? ? ? ?Retry ?long limit:7 ? RTS thr:off ? Fragment thr:off
> ? ? ? ? ? ? ?Encryption key:off
> ? ? ? ? ? ? ?Power Management:off
> ? ? ? ? ? ? ?Link Quality=70/70 ?Signal level=-29 dBm
> ? ? ? ? ? ? ?Rx invalid nwid:0 ?Rx invalid crypt:0 ?Rx invalid frag:0
> ? ? ? ? ? ? ?Tx excessive retries:2 ?Invalid misc:90 ? Missed beacon:0
>
> All my laptops run Debian unstable, currently with kernel 3.1~rc7 (from
> Debian).
>
> I have already tested changing the security scheme from WPA (AES) to WPA
> (AES/TKIP), WPA (TKIP), WEP and even without security, but the problem is
> the same.
>
> I own other laptops using the brcmsmac and b43 modules which do not have any
> problem with this router. On the other hand, I have successfully tested the
> laptop chip with other routers with success.
>
> I normally use NetworkManager as a wireless client, but have also tested
> wicd and manual configuration (through Debian ifupdown and wpa_supplicant),
> but still encounter the same problem.
>
> I do not have much control on the router as it is the one supplied by my
> internet provider, hence I cannot access to the logs.
>
> I suspect it might be a problem of the ath9k module and/or in the router
> itself, but I do not know how to confirm any of these speculations.
>
> I have already noticed similar issues [0] in the archive of the list, which
> lets me think I am in the right direction, but what extra information would
> you need to help you understand the issue?

Hi Julien,

i can allocate some of my free time and try to recreate the issue here
next week. may be a regression.
enabling mac80211 debugs will also be useful. you said that the issue
happens without security, and  also only when the AP router is already
connected by another station. in the other thread it was mentioned
that the problem occurs only when its a wireless router.
AR9285 has diversity enabled, we need to check if its working properly.
also please check whether setting coverage class via iw'improves the situation.
iw phy <phyname> set coverage <coverage class>


>
> Cheers,
> Julien
>
> [0] https://lists.ath9k.org/pipermail/ath9k-devel/2011-October/007055.html
>
> --
> ?.''`. ? Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>
> ?: :' ?: ?Debian Developer & Free software contributor
> ?`. `'` ? http://www.kirya.net/
> ? `- ? ? 4096R/ E1D8 5796 8214 4687 E416 ?948C 859F EF67 258E 26B1
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>
>



-- 
shafi

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Hi Mohammed,

Le mardi 04 oct. 2011 ? 07:46:58 (+0200 CEST), Mohammed Shafi a ?crit?:
> On Mon, Oct 3, 2011 at 10:06 PM, Julien Valroff <julien@kirya.net> wrote:
> > [ please CC me when replying, I am not subscribed to the list ]
> >
> > Hi,
> >
> > I have recently purchased a new laptop (Asus X93S) with an Atheros AR9285
> > wireless chip which obviously uses the ath9k module.
> >
> > When connecting to the AP of my router, everything works as expected
> > provided no other PC has already been connected to the AP. If that is not
> > the case, I have to reboot the router and make sure my new laptop is the
> > first to connect.
[...]
> i can allocate some of my free time and try to recreate the issue here
> next week. may be a regression.

Thanks for taking the time to have a look at this issue.

> enabling mac80211 debugs will also be useful. you said that the issue
> happens without security, and  also only when the AP router is already
> connected by another station. in the other thread it was mentioned
> that the problem occurs only when its a wireless router.

I am compiling a kernel with mac80211 debug features enabled, will make
further tests within the next week-end and report.

> AR9285 has diversity enabled, we need to check if its working properly.
> also please check whether setting coverage class via iw'improves the situation.
> iw phy <phyname> set coverage <coverage class>

This doesn't help - I've tried using set distance 0 (default), and 5 (my AP
is less than 1m from my laptop).

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le mardi 04 oct. 2011 ? 20:35:14 (+0200 CEST), Julien Valroff a ?crit?:
> Hi Mohammed,
> 
> Le mardi 04 oct. 2011 ? 07:46:58 (+0200 CEST), Mohammed Shafi a ?crit?:
> > On Mon, Oct 3, 2011 at 10:06 PM, Julien Valroff <julien@kirya.net> wrote:
> > > [ please CC me when replying, I am not subscribed to the list ]
> > >
> > > Hi,
> > >
> > > I have recently purchased a new laptop (Asus X93S) with an Atheros AR9285
> > > wireless chip which obviously uses the ath9k module.
> > >
> > > When connecting to the AP of my router, everything works as expected
> > > provided no other PC has already been connected to the AP. If that is not
> > > the case, I have to reboot the router and make sure my new laptop is the
> > > first to connect.
> [...]
[...]
> > enabling mac80211 debugs will also be useful. you said that the issue
> > happens without security, and  also only when the AP router is already
> > connected by another station. in the other thread it was mentioned
> > that the problem occurs only when its a wireless router.
> 
> I am compiling a kernel with mac80211 debug features enabled, will make
> further tests within the next week-end and report.

Please find attached the kernel log with mac80211 debug enabled.

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1
-------------- next part --------------
Oct  5 06:11:42 ares kernel: [   18.571094] ieee80211 phy0: device now idle
Oct  5 06:11:42 ares kernel: [   18.580945] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:42 ares kernel: [   18.592499] wlan0: authenticate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:42 ares kernel: [   18.594809] wlan0: authenticated
Oct  5 06:11:42 ares kernel: [   18.594836] ieee80211 phy0: device now idle
Oct  5 06:11:42 ares kernel: [   18.614520] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:42 ares kernel: [   18.625980] wlan0: associate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:42 ares kernel: [   18.636510] wlan0: RX AssocResp from 9a:5d:6e:56:24:8c (capab=0x411 status=0 aid=2)
Oct  5 06:11:42 ares kernel: [   18.636514] wlan0: associated
Oct  5 06:11:42 ares kernel: [   18.636522] ieee80211 phy0: Allocated STA 9a:5d:6e:56:24:8c
Oct  5 06:11:42 ares kernel: [   18.636547] ieee80211 phy0: Inserted STA 9a:5d:6e:56:24:8c
Oct  5 06:11:42 ares kernel: [   18.636551] ieee80211 phy0: WMM queue=2 aci=0 acm=0 aifs=3 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:42 ares kernel: [   18.636577] ieee80211 phy0: WMM queue=3 aci=1 acm=0 aifs=7 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:42 ares kernel: [   18.636598] ieee80211 phy0: WMM queue=1 aci=2 acm=0 aifs=2 cWmin=7 cWmax=15 txop=94 uapsd=0
Oct  5 06:11:42 ares kernel: [   18.636621] ieee80211 phy0: WMM queue=0 aci=3 acm=0 aifs=2 cWmin=3 cWmax=7 txop=47 uapsd=0
Oct  5 06:11:42 ares kernel: [   18.644169] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
Oct  5 06:11:42 ares kernel: [   18.652046] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:43 ares kernel: [   19.368042] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.368061] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.376046] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.496031] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.658929] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.725231] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:44 ares kernel: [   20.909941] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:45 ares kernel: [   21.160704] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:45 ares kernel: [   21.361498] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:48 ares kernel: [   24.206195] net_ratelimit: 2 callbacks suppressed
Oct  5 06:11:48 ares kernel: [   24.206199] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:48 ares kernel: [   24.376046] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:48 ares kernel: [   24.843389] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:49 ares kernel: [   24.997908] wlan0: deauthenticated from 9a:5d:6e:56:24:8c (Reason: 15)
Oct  5 06:11:49 ares kernel: [   25.068029] ieee80211 phy0: Removed STA 9a:5d:6e:56:24:8c
Oct  5 06:11:49 ares kernel: [   25.068042] ieee80211 phy0: Destroyed STA 9a:5d:6e:56:24:8c
Oct  5 06:11:49 ares kernel: [   25.068083] ieee80211 phy0: device now idle
Oct  5 06:11:49 ares kernel: [   25.075116] cfg80211: Calling CRDA to update world regulatory domain
Oct  5 06:11:49 ares kernel: [   25.179073] ieee80211 phy0: device no longer idle - scanning
Oct  5 06:11:50 ares kernel: [   26.219091] ieee80211 phy0: device now idle
Oct  5 06:11:50 ares kernel: [   26.228908] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:50 ares kernel: [   26.239454] wlan0: authenticate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:50 ares kernel: [   26.241529] wlan0: authenticated
Oct  5 06:11:50 ares kernel: [   26.241554] ieee80211 phy0: device now idle
Oct  5 06:11:50 ares kernel: [   26.251479] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:50 ares kernel: [   26.262037] wlan0: associate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:50 ares kernel: [   26.271195] wlan0: RX ReassocResp from 9a:5d:6e:56:24:8c (capab=0x411 status=0 aid=2)
Oct  5 06:11:50 ares kernel: [   26.271199] wlan0: associated
Oct  5 06:11:50 ares kernel: [   26.271206] ieee80211 phy0: Allocated STA 9a:5d:6e:56:24:8c
Oct  5 06:11:50 ares kernel: [   26.271227] ieee80211 phy0: Inserted STA 9a:5d:6e:56:24:8c
Oct  5 06:11:50 ares kernel: [   26.271233] ieee80211 phy0: WMM queue=2 aci=0 acm=0 aifs=3 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:50 ares kernel: [   26.271256] ieee80211 phy0: WMM queue=3 aci=1 acm=0 aifs=7 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:50 ares kernel: [   26.271280] ieee80211 phy0: WMM queue=1 aci=2 acm=0 aifs=2 cWmin=7 cWmax=15 txop=94 uapsd=0
Oct  5 06:11:50 ares kernel: [   26.271304] ieee80211 phy0: WMM queue=0 aci=3 acm=0 aifs=2 cWmin=3 cWmax=7 txop=47 uapsd=0
Oct  5 06:11:52 ares kernel: [   28.084027] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:52 ares kernel: [   28.384035] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:53 ares kernel: [   29.085280] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:53 ares kernel: [   29.384024] wlan0: no IPv6 routers present
Oct  5 06:11:55 ares kernel: [   31.087560] wlan0: dropped frame to 9a:5d:6e:56:24:8c (unauthorized port)
Oct  5 06:11:57 ares kernel: [   33.004957] wlan0: deauthenticated from 9a:5d:6e:56:24:8c (Reason: 15)
Oct  5 06:11:57 ares kernel: [   33.052038] ieee80211 phy0: Removed STA 9a:5d:6e:56:24:8c
Oct  5 06:11:57 ares kernel: [   33.052064] ieee80211 phy0: Destroyed STA 9a:5d:6e:56:24:8c
Oct  5 06:11:57 ares kernel: [   33.052090] ieee80211 phy0: device now idle
Oct  5 06:11:57 ares kernel: [   33.059080] cfg80211: Calling CRDA to update world regulatory domain
Oct  5 06:11:57 ares kernel: [   33.162986] ieee80211 phy0: device no longer idle - scanning
Oct  5 06:11:58 ares kernel: [   34.199084] ieee80211 phy0: device now idle
Oct  5 06:11:58 ares kernel: [   34.208888] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:58 ares kernel: [   34.219399] wlan0: authenticate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:58 ares kernel: [   34.221379] wlan0: authenticated
Oct  5 06:11:58 ares kernel: [   34.221407] ieee80211 phy0: device now idle
Oct  5 06:11:58 ares kernel: [   34.231343] ieee80211 phy0: device no longer idle - working
Oct  5 06:11:58 ares kernel: [   34.241880] wlan0: associate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:11:58 ares kernel: [   34.251090] wlan0: RX ReassocResp from 9a:5d:6e:56:24:8c (capab=0x411 status=0 aid=2)
Oct  5 06:11:58 ares kernel: [   34.251094] wlan0: associated
Oct  5 06:11:58 ares kernel: [   34.251102] ieee80211 phy0: Allocated STA 9a:5d:6e:56:24:8c
Oct  5 06:11:58 ares kernel: [   34.251122] ieee80211 phy0: Inserted STA 9a:5d:6e:56:24:8c
Oct  5 06:11:58 ares kernel: [   34.251129] ieee80211 phy0: WMM queue=2 aci=0 acm=0 aifs=3 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:58 ares kernel: [   34.251160] ieee80211 phy0: WMM queue=3 aci=1 acm=0 aifs=7 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:11:58 ares kernel: [   34.251184] ieee80211 phy0: WMM queue=1 aci=2 acm=0 aifs=2 cWmin=7 cWmax=15 txop=94 uapsd=0
Oct  5 06:11:58 ares kernel: [   34.251207] ieee80211 phy0: WMM queue=0 aci=3 acm=0 aifs=2 cWmin=3 cWmax=7 txop=47 uapsd=0
Oct  5 06:12:05 ares kernel: [   41.011922] wlan0: deauthenticated from 9a:5d:6e:56:24:8c (Reason: 15)
Oct  5 06:12:05 ares kernel: [   41.068070] ieee80211 phy0: Removed STA 9a:5d:6e:56:24:8c
Oct  5 06:12:05 ares kernel: [   41.068083] ieee80211 phy0: Destroyed STA 9a:5d:6e:56:24:8c
Oct  5 06:12:05 ares kernel: [   41.068109] ieee80211 phy0: device now idle
Oct  5 06:12:05 ares kernel: [   41.075094] cfg80211: Calling CRDA to update world regulatory domain
Oct  5 06:12:05 ares kernel: [   41.179027] ieee80211 phy0: device no longer idle - scanning
Oct  5 06:12:06 ares kernel: [   42.215125] ieee80211 phy0: device now idle
Oct  5 06:12:06 ares kernel: [   42.224941] ieee80211 phy0: device no longer idle - working
Oct  5 06:12:06 ares kernel: [   42.235497] wlan0: authenticate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:12:06 ares kernel: [   42.237503] wlan0: authenticated
Oct  5 06:12:06 ares kernel: [   42.237531] ieee80211 phy0: device now idle
Oct  5 06:12:06 ares kernel: [   42.247464] ieee80211 phy0: device no longer idle - working
Oct  5 06:12:06 ares kernel: [   42.258025] wlan0: associate with 9a:5d:6e:56:24:8c (try 1)
Oct  5 06:12:06 ares kernel: [   42.267216] wlan0: RX ReassocResp from 9a:5d:6e:56:24:8c (capab=0x411 status=0 aid=2)
Oct  5 06:12:06 ares kernel: [   42.267220] wlan0: associated
Oct  5 06:12:06 ares kernel: [   42.267227] ieee80211 phy0: Allocated STA 9a:5d:6e:56:24:8c
Oct  5 06:12:06 ares kernel: [   42.267247] ieee80211 phy0: Inserted STA 9a:5d:6e:56:24:8c
Oct  5 06:12:06 ares kernel: [   42.267251] ieee80211 phy0: WMM queue=2 aci=0 acm=0 aifs=3 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:12:06 ares kernel: [   42.267273] ieee80211 phy0: WMM queue=3 aci=1 acm=0 aifs=7 cWmin=15 cWmax=1023 txop=0 uapsd=0
Oct  5 06:12:06 ares kernel: [   42.267295] ieee80211 phy0: WMM queue=1 aci=2 acm=0 aifs=2 cWmin=7 cWmax=15 txop=94 uapsd=0
Oct  5 06:12:06 ares kernel: [   42.267317] ieee80211 phy0: WMM queue=0 aci=3 acm=0 aifs=2 cWmin=3 cWmax=7 txop=47 uapsd=0
Oct  5 06:12:07 ares kernel: [   43.008681] wlan0: deauthenticating from 9a:5d:6e:56:24:8c by local choice (reason=3)
Oct  5 06:12:07 ares kernel: [   43.100038] ieee80211 phy0: Removed STA 9a:5d:6e:56:24:8c
Oct  5 06:12:07 ares kernel: [   43.100063] ieee80211 phy0: Destroyed STA 9a:5d:6e:56:24:8c
Oct  5 06:12:07 ares kernel: [   43.100088] ieee80211 phy0: device now idle
Oct  5 06:12:07 ares kernel: [   43.107074] cfg80211: Calling CRDA to update world regulatory domain
Oct  5 06:12:30 ares kernel: [   66.006708] ieee80211 phy0: device no longer idle - scanning
Oct  5 06:12:31 ares kernel: [   67.791088] ieee80211 phy0: device now idle

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Hi Mohammed,

Le mercredi 05 oct. 2011 ? 06:24:06 (+0200 CEST), Julien Valroff a ?crit?:
> Le mardi 04 oct. 2011 ? 20:35:14 (+0200 CEST), Julien Valroff a ?crit?:
> > Le mardi 04 oct. 2011 ? 07:46:58 (+0200 CEST), Mohammed Shafi a ?crit?:
[...]
> > > enabling mac80211 debugs will also be useful. you said that the issue
> > > happens without security, and  also only when the AP router is already
> > > connected by another station. in the other thread it was mentioned
> > > that the problem occurs only when its a wireless router.
> > 
> > I am compiling a kernel with mac80211 debug features enabled, will make
> > further tests within the next week-end and report.
> 
> Please find attached the kernel log with mac80211 debug enabled.

Have you been able to reproduce the issue on your side?

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Mohammed Shafi]

On Sat, Oct 15, 2011 at 10:21 AM, Julien Valroff <julien@kirya.net> wrote:
> Hi Mohammed,
>
> Le mercredi 05 oct. 2011 ? 06:24:06 (+0200 CEST), Julien Valroff a ?crit?:
>> Le mardi 04 oct. 2011 ? 20:35:14 (+0200 CEST), Julien Valroff a ?crit?:
>> > Le mardi 04 oct. 2011 ? 07:46:58 (+0200 CEST), Mohammed Shafi a ?crit?:
> [...]
>> > > enabling mac80211 debugs will also be useful. you said that the issue
>> > > happens without security, and ?also only when the AP router is already
>> > > connected by another station. in the other thread it was mentioned
>> > > that the problem occurs only when its a wireless router.
>> >
>> > I am compiling a kernel with mac80211 debug features enabled, will make
>> > further tests within the next week-end and report.
>>
>> Please find attached the kernel log with mac80211 debug enabled.
>
> Have you been able to reproduce the issue on your side?

sorry no :(, i used compat-wireless-2011-10-10  and i made sure that
my ar9285 is to be connected to an AP which is already  connected by
another station. i did not get the issue. the environment is a less a
nosier one and the channel is 7 same as yours.

just doubt, why your tx-power is only 14dbm while mine has 27 dbm in
iwconfig. i assume its just a random sample.  if its consistently
14dbm when you connect your AR9285 station first, we need to look at
that.

>
> Cheers,
> Julien
>
> --
> ?.''`. ? Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>
> ?: :' ?: ?Debian Developer & Free software contributor
> ?`. `'` ? http://www.kirya.net/
> ? `- ? ? 4096R/ E1D8 5796 8214 4687 E416 ?948C 859F EF67 258E 26B1
>



-- 
shafi

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le dimanche 16 oct. 2011 ? 15:22:45 (+0200 CEST), Mohammed Shafi a ?crit?:
> On Sat, Oct 15, 2011 at 10:21 AM, Julien Valroff <julien@kirya.net> wrote:
> > Hi Mohammed,
> >
> > Le mercredi 05 oct. 2011 ? 06:24:06 (+0200 CEST), Julien Valroff a ?crit?:
> >> Le mardi 04 oct. 2011 ? 20:35:14 (+0200 CEST), Julien Valroff a ?crit?:
> >> > Le mardi 04 oct. 2011 ? 07:46:58 (+0200 CEST), Mohammed Shafi a ?crit?:
> > [...]
> >> > > enabling mac80211 debugs will also be useful. you said that the issue
> >> > > happens without security, and ?also only when the AP router is already
> >> > > connected by another station. in the other thread it was mentioned
> >> > > that the problem occurs only when its a wireless router.
> >> >
> >> > I am compiling a kernel with mac80211 debug features enabled, will make
> >> > further tests within the next week-end and report.
> >>
> >> Please find attached the kernel log with mac80211 debug enabled.
> >
> > Have you been able to reproduce the issue on your side?
> 
> sorry no :(, i used compat-wireless-2011-10-10  and i made sure that
> my ar9285 is to be connected to an AP which is already  connected by
> another station. i did not get the issue. the environment is a less a
> nosier one and the channel is 7 same as yours.

The channel doesn't matter in my case.

> just doubt, why your tx-power is only 14dbm while mine has 27 dbm in
> iwconfig. i assume its just a random sample.  if its consistently
> 14dbm when you connect your AR9285 station first, we need to look at
> that.

It is always 14dbm. Trying to change it manually to a higher value fails:
Error for wireless request "Set Tx Power" (8B26) :
    SET failed on device wlan0 ; Invalid argument.

All my other laptops have 19dBm.

I have also just noticed a new symptom: in order to check the tx-power
value, I have disconnected my other laptops from the AP, rebooted the AP. I
was then able to connect to my AP. When re-connecting another laptop to the
AP, the AR9285 laptop has dropped the connection, and is now unable to
connect again to the AP (until the next AP reboot).

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Adrian Chadd]

Can you please grab a third device and grab a dump of the actual
probe/association frames?

I think it'd be helpful to see what's changed.



Adrian

[ath9k-devel] Can't associate with a particular AP

[Mohammed Shafi]

On Sun, Oct 16, 2011 at 9:01 PM, Adrian Chadd <adrian@freebsd.org> wrote:
> Can you please grab a third device and grab a dump of the actual
> probe/association frames?

looks like the tx-power is constrained due to regulatory (or) power
constraint element from AP?

>
> I think it'd be helpful to see what's changed.
>
>
>
> Adrian
>



-- 
shafi

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le dimanche 16 oct. 2011 ? 17:31:44 (+0200 CEST), Adrian Chadd a ?crit?:
> Can you please grab a third device and grab a dump of the actual
> probe/association frames?
> 
> I think it'd be helpful to see what's changed.

Would you please explain in details what you want me to do? Not sure to
understand your question...

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> Le dimanche 16 oct. 2011 ? 17:31:44 (+0200 CEST), Adrian Chadd a ?crit?:
> > Can you please grab a third device and grab a dump of the actual
> > probe/association frames?
> > 
> > I think it'd be helpful to see what's changed.
> 
> Would you please explain in details what you want me to do? Not sure
> to understand your question...

Use another computer with a wireless card that supports monitoring
mode. Perhaps the intel card. Disable everything that is clickety
clickety (wpa-supplicant, Network Manager, wicd, everything like
that) and run:

ip l s dev eth1 down
iwconfig eth1 mode monitoring
ip l s dev eth1 up
wireshark

Select to listen on eth1 in wireshark (assuming this is your intel
wifi) and capture the packets from both a successful association by
ath9k, as well as an unsuccessful on. Save the captures, and share
them with Adrian and/or the list, if they aren't very big.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Hi,

Le dimanche 16 oct. 2011 ? 17:31:44 (+0200 CEST), Adrian Chadd a ?crit?:
> Can you please grab a third device and grab a dump of the actual
> probe/association frames?
> 
> I think it'd be helpful to see what's changed.

I have tried and captured the association process from my 2 other laptops
using wireshark, but unfortunately, I am unable to capture packets in
monitoring mode: in Wireshark, the box is not greyed out, but I cannot tick
it.

I've tried reading the Wireshark documentation on this point, but I could
not see anything about this problem.

I have followed Peter's instructions and the ones described in [0] both
with the same results.

But still, I have been able to capture something from the problematic
laptop, you will find the wireshark logs attached.

When failing to associate, I can see a lot of lines as follows:
50	11.081661	ba:f4:8a:c6:c8:44 (TA)	74:2f:68:57:2c:dc (RA)	802.11	46	Request-to-send, Flags=........C

followed by:
166	23.082827	ba:f4:8a:c6:c8:44	74:2f:68:57:2c:dc	802.11	56	Deauthentication, SN=1142, FN=0, Flags=........C

where:
ba:f4:8a:c6:c8:44 is my AP
74:2f:68:57:2c:dc is my Atheros wlan NIC

When the association succeeds, I notice the MAC address of the AP changes:
99	5.752088	c6:6e:27:a6:be:a4	74:2f:68:57:2c:dc	802.11	272	Probe Response, SN=269, FN=0, Flags=........C, BI=100, SSID=Kirya
                      ^^^^

Which is confirmed by iwlist scan.
Before the association:
Cell 01 - Address: BA:F4:8A:C6:C8:44
        Channel:3
        Frequency:2.422 GHz (Channel 3)
        Quality=26/70  Signal level=-84 dBm  
        Encryption key:on
        ESSID:"Kirya"

After the succeeded association:
Cell 04 - Address: C6:6E:27:A6:BE:A4
        Channel:3
        Frequency:2.422 GHz (Channel 3)
        Quality=70/70  Signal level=-35 dBm  
        Encryption key:on
        ESSID:"Kirya"

How is this possible?

Please continue and CC me as I am not subscribed to the list.

Cheers,
Julien

[0] http://wiki.wireshark.org/CaptureSetup/WLAN#Linux

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wireshark_failed-association
Type: application/octet-stream
Size: 27710 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111023/cb08a59e/attachment-0002.obj 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wireshark_succeeded-association
Type: application/octet-stream
Size: 46875 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111023/cb08a59e/attachment-0003.obj 

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> I have tried and captured the association process from my 2 other laptops
> using wireshark, but unfortunately, I am unable to capture packets in
> monitoring mode: in Wireshark, the box is not greyed out, but I cannot tick
> it.

I think it is important to iwconfig eth1 mode monitor before you run
ip link set dev eth1 up, and once you have done that it does not
matter what settings wireshark does or does not do, the mode can't be
changed when the interface is up, so just starting the capture should
work.


> When the association succeeds, I notice the MAC address of the AP
> changes:
..
> Before the association:
> Cell 01 - Address: BA:F4:8A:C6:C8:44
..
> After the succeeded association:
> Cell 04 - Address: C6:6E:27:A6:BE:A4

Which is the actual correct address of your AP? I would expect
c6-6e-27, but the IEEE OUI listing does not have a match for it!

http://standards.ieee.org/cgi-bin/ouisearch?c6-6e-27


> How is this possible?

I guess some fundamental data corruption problem in ath9k. I'm not
surprised at all, although developers claim that by now the code
quality is good. This is a total BS error. I could rant for forever.


> Please continue and CC me as I am not subscribed to the list.

I would try FreeBSD on that machine. I believe the driver there is
more reliable, and using it can allow you to get more concrete help
from Adrian in case the problem is seen also there.


//Peter

[ath9k-devel] Can't associate with a particular AP

[kang haiyang]

from the failed wireshark log, i saw there is inconsistent information 
in reassociation response:
in capability tag: ap tells client that it supports 40Mhz, but in the HT 
info tags it tells client that it only support 20Mhz.

i don't know if this maybe bother some particular clients.

Best Regards,
kang haiyang


On 10/23/2011 10:25 PM, Julien Valroff wrote:
> Hi,
>
> Le dimanche 16 oct. 2011 ? 17:31:44 (+0200 CEST), Adrian Chadd a ?crit :
>> Can you please grab a third device and grab a dump of the actual
>> probe/association frames?
>>
>> I think it'd be helpful to see what's changed.
> I have tried and captured the association process from my 2 other laptops
> using wireshark, but unfortunately, I am unable to capture packets in
> monitoring mode: in Wireshark, the box is not greyed out, but I cannot tick
> it.
>
> I've tried reading the Wireshark documentation on this point, but I could
> not see anything about this problem.
>
> I have followed Peter's instructions and the ones described in [0] both
> with the same results.
>
> But still, I have been able to capture something from the problematic
> laptop, you will find the wireshark logs attached.
>
> When failing to associate, I can see a lot of lines as follows:
> 50	11.081661	ba:f4:8a:c6:c8:44 (TA)	74:2f:68:57:2c:dc (RA)	802.11	46	Request-to-send, Flags=........C
>
> followed by:
> 166	23.082827	ba:f4:8a:c6:c8:44	74:2f:68:57:2c:dc	802.11	56	Deauthentication, SN=1142, FN=0, Flags=........C
>
> where:
> ba:f4:8a:c6:c8:44 is my AP
> 74:2f:68:57:2c:dc is my Atheros wlan NIC
>
> When the association succeeds, I notice the MAC address of the AP changes:
> 99	5.752088	c6:6e:27:a6:be:a4	74:2f:68:57:2c:dc	802.11	272	Probe Response, SN=269, FN=0, Flags=........C, BI=100, SSID=Kirya
>                        ^^^^
>
> Which is confirmed by iwlist scan.
> Before the association:
> Cell 01 - Address: BA:F4:8A:C6:C8:44
>          Channel:3
>          Frequency:2.422 GHz (Channel 3)
>          Quality=26/70  Signal level=-84 dBm
>          Encryption key:on
>          ESSID:"Kirya"
>
> After the succeeded association:
> Cell 04 - Address: C6:6E:27:A6:BE:A4
>          Channel:3
>          Frequency:2.422 GHz (Channel 3)
>          Quality=70/70  Signal level=-35 dBm
>          Encryption key:on
>          ESSID:"Kirya"
>
> How is this possible?
>
> Please continue and CC me as I am not subscribed to the list.
>
> Cheers,
> Julien
>
> [0] http://wiki.wireshark.org/CaptureSetup/WLAN#Linux
>
>
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111023/942575eb/attachment.htm 

[ath9k-devel] Can't associate with a particular AP

[Adrian Chadd]

I'm kind of surprised that you're seeing different AP MACs, where did
you take that capture from?



Adrian

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le dimanche 23 oct. 2011 ? 17:54:07 (+0200 CEST), Adrian Chadd a ?crit?:
> I'm kind of surprised that you're seeing different AP MACs, where did
> you take that capture from?

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le dimanche 23 oct. 2011 ? 17:02:40 (+0200 CEST), Peter Stuge a ?crit?:
> Julien Valroff wrote:
> > I have tried and captured the association process from my 2 other laptops
> > using wireshark, but unfortunately, I am unable to capture packets in
> > monitoring mode: in Wireshark, the box is not greyed out, but I cannot tick
> > it.
> 
> I think it is important to iwconfig eth1 mode monitor before you run
> ip link set dev eth1 up, and once you have done that it does not
> matter what settings wireshark does or does not do, the mode can't be
> changed when the interface is up, so just starting the capture should
> work.

I couldn't capture anything this way from the other laptops.

As explained, I have been able to setup a mon0 interface on the laptop with
the ath9k module.

> > When the association succeeds, I notice the MAC address of the AP
> > changes:
> ..
> > Before the association:
> > Cell 01 - Address: BA:F4:8A:C6:C8:44
> ..
> > After the succeeded association:
> > Cell 04 - Address: C6:6E:27:A6:BE:A4
> 
> Which is the actual correct address of your AP? I would expect
> c6-6e-27, but the IEEE OUI listing does not have a match for it!

I would also say that, given what I can see from the other laptops.

> > How is this possible?
> 
> I guess some fundamental data corruption problem in ath9k. I'm not
> surprised at all, although developers claim that by now the code
> quality is good. This is a total BS error. I could rant for forever.

Are you sure it could not be a problem with the AP? Or an incompatibility
between both, as I have no problem with the same AP with other
laptops/smartphones.

> > Please continue and CC me as I am not subscribed to the list.
> 
> I would try FreeBSD on that machine. I believe the driver there is
> more reliable, and using it can allow you to get more concrete help
> from Adrian in case the problem is seen also there.

I will try something like a livecd, I do not want to install FreeBSD.
Or it could be a good opportunity to test Debian GNU/kFreeBSD ;)

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> > I think it is important to iwconfig eth1 mode monitor before you run
> > ip link set dev eth1 up, and once you have done that it does not
> > matter what settings wireshark does or does not do, the mode can't be
> > changed when the interface is up, so just starting the capture should
> > work.
> 
> I couldn't capture anything this way from the other laptops.

How does it fail?


> > > Cell 01 - Address: BA:F4:8A:C6:C8:44
> > ..
> > > Cell 04 - Address: C6:6E:27:A6:BE:A4
> > 
> > Which is the actual correct address of your AP? I would expect
> > c6-6e-27, but the IEEE OUI listing does not have a match for it!
> 
> I would also say that, given what I can see from the other laptops.

I would check the sticker on the AP.


> > > How is this possible?
> > 
> > I guess some fundamental data corruption problem in ath9k. I'm not
> > surprised at all, although developers claim that by now the code
> > quality is good. This is a total BS error. I could rant for forever.
> 
> Are you sure it could not be a problem with the AP?

If ath9k can associate sometimes it really must be able to associate
every time, especially since other drivers can.


> Or an incompatibility between both,

This would of course be unacceptable, since ath9k is supposed to
implement a standard, the purpose of which is interoperability.


> as I have no problem with the same AP with other
> laptops/smartphones.

The other laptops likely have different hardware and software. If you
want you can gather more data by comparing results while you
eliminate those differences one at a time.


> > I would try FreeBSD on that machine. I believe the driver there is
> > more reliable, and using it can allow you to get more concrete help
> > from Adrian in case the problem is seen also there.
> 
> I will try something like a livecd,

Might work, but by doing that you are only adding more parameters,
which is absolutely counter-productive for debugging.


> I do not want to install FreeBSD.

This is stupid. Either you want to resolve the problem or you do not.
If you do not then you would not have posted to the list. If
installing FreeBSD can help identify the problem then that is
obviously what you should do.


> Or it could be a good opportunity to test Debian GNU/kFreeBSD ;)

Again a really bad idea. The purpose of using FreeBSD would be to
get another data point, and to take advantage of Adrian's expertise.

The only way of doing that which makes sense is to use exactly the
same code that Adrian has; which means you should build the very
latest fbsd world and kernel, and see what happens. This should be
really easy given that you have experience from distribution
development already. Also, this should be absolutely obvious to you
already. I'm surprised I have to explain it.

It's simple: Either you want to try to identify the problem, or you
don't care so much. If you don't care then you can't expect anyone
else to, and the problem will stay unresolved. You have an
opportunity to improve ath9k for the benefit of all, and you do not
want to do it because you would have to install FreeBSD? lolwut?


//Peter

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le dimanche 23 oct. 2011 ? 18:52:07 (+0200 CEST), Peter Stuge a ?crit?:
> Julien Valroff wrote:
> > > I think it is important to iwconfig eth1 mode monitor before you run
> > > ip link set dev eth1 up, and once you have done that it does not
> > > matter what settings wireshark does or does not do, the mode can't be
> > > changed when the interface is up, so just starting the capture should
> > > work.
> > 
> > I couldn't capture anything this way from the other laptops.
> 
> How does it fail?

No error, simply no packet being capture.

> > > > Cell 01 - Address: BA:F4:8A:C6:C8:44
> > > ..
> > > > Cell 04 - Address: C6:6E:27:A6:BE:A4
> > > 
> > > Which is the actual correct address of your AP? I would expect
> > > c6-6e-27, but the IEEE OUI listing does not have a match for it!
> > 
> > I would also say that, given what I can see from the other laptops.
> 
> I would check the sticker on the AP.

The only MAC being stated on the sticker is the one for the ethernet
interface (as explained, it is the router of my internet provider).

[...]

> > as I have no problem with the same AP with other
> > laptops/smartphones.
> 
> The other laptops likely have different hardware and software. If you
> want you can gather more data by comparing results while you
> eliminate those differences one at a time.

I'll do this.

> > > I would try FreeBSD on that machine. I believe the driver there is
> > > more reliable, and using it can allow you to get more concrete help
> > > from Adrian in case the problem is seen also there.
> > 
> > I will try something like a livecd,
> 
> Might work, but by doing that you are only adding more parameters,
> which is absolutely counter-productive for debugging.

That's just a more convenient way to start.
 
> > I do not want to install FreeBSD.
> 
> This is stupid. Either you want to resolve the problem or you do not.
> If you do not then you would not have posted to the list. If
> installing FreeBSD can help identify the problem then that is
> obviously what you should do.

I can do that, but this will have to wait - as I'd like to supply as many
details as possible, I'd thought that 

> > Or it could be a good opportunity to test Debian GNU/kFreeBSD ;)
> 
> Again a really bad idea. The purpose of using FreeBSD would be to
> get another data point, and to take advantage of Adrian's expertise.

See the smiley at the end of my sentence

> The only way of doing that which makes sense is to use exactly the
> same code that Adrian has; which means you should build the very
> latest fbsd world and kernel, and see what happens. This should be
> really easy given that you have experience from distribution
> development already. Also, this should be absolutely obvious to you
> already. I'm surprised I have to explain it.
> 
> It's simple: Either you want to try to identify the problem, or you
> don't care so much. If you don't care then you can't expect anyone
> else to, and the problem will stay unresolved. You have an
> opportunity to improve ath9k for the benefit of all, and you do not
> want to do it because you would have to install FreeBSD? lolwut?

I have absolutely *no* experience with FreeBSD, and we are talking about my
main laptop, the data of which I need everyday. Before installing FreeBSD, I
need to move 500Gb to another computer which I do not have on hand, this
explains why I am reluctant to install FreeBSD /now/.

Sorry if I haven't understood your initial request, I had considered it as a
"give it a try and see if it works"?

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> > > > I think it is important to iwconfig eth1 mode monitor before you run
> > > > ip link set dev eth1 up, and once you have done that it does not
> > > > matter what settings wireshark does or does not do, the mode can't be
> > > > changed when the interface is up, so just starting the capture should
> > > > work.
> > > 
> > > I couldn't capture anything this way from the other laptops.
> > 
> > How does it fail?
> 
> No error, simply no packet being capture.

Double check that the channel is set to what your AP uses?


> > > > Which is the actual correct address of your AP? I would expect
> > > > c6-6e-27, but the IEEE OUI listing does not have a match for it!
> > > 
> > > I would also say that, given what I can see from the other laptops.
> > 
> > I would check the sticker on the AP.
> 
> The only MAC being stated on the sticker is the one for the ethernet
> interface (as explained, it is the router of my internet provider).

I see. What's the first three bytes of the ethernet MAC?


> > The other laptops likely have different hardware and software. If you
> > want you can gather more data by comparing results while you
> > eliminate those differences one at a time.
> 
> I'll do this.

It's lots of work, but ok..


> > > I do not want to install FreeBSD.
> > 
> > This is stupid. Either you want to resolve the problem or you do not.
> > If you do not then you would not have posted to the list. If
> > installing FreeBSD can help identify the problem then that is
> > obviously what you should do.
> 
> I can do that, but this will have to wait -

It'll take maybe an evening. Recommend #freebsd for help, but the
documentation is already really good. Key word: you want -CURRENT.


> > you do not want to do it because you would have to install
> > FreeBSD? lolwut?
> 
> I have absolutely *no* experience with FreeBSD,

I didn't either. But it's really not strange.


> and we are talking about my main laptop, the data of which I need
> everyday. Before installing FreeBSD, I need to move 500Gb to
> another computer which I do not have on hand, this explains why I
> am reluctant to install FreeBSD /now/.

Um, I didn't say that you should convert your system and use FreeBSD
forever. In this case it would be a tool to get a very significant
data point, and the only real way to get help from a skilled driver
developer, since Atheros insist on trying to reproduce issues, which
is obviously useless for a case like yours.

Swap harddrives or even install to a USB stick.. Don't start messing
with your existing system.


> Sorry if I haven't understood your initial request, I had
> considered it as a "give it a try and see if it works"?

It was meant more firmly than that. The point is that Adrian tries to
work with remote problems.

If all works perfectly in fbsd then it is confirmed that you are
seeing an ath9k software error. This may perhaps also motivate
Atheros more. Currently your problem is just too unclear, and in a
bunch of other very unclear problems.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Adrian Chadd]

Just try a PCBSD USB image, which should fire up X as part of the installer.
I think the pcbsd usb images (full and bootonly) are "live", as in
give you a basic desktop environment.

You _should_ be able to configure the wireless through the installer
environment.
That should be enough to verify whether the wifi is working or not.


Adrian

[ath9k-devel] Can't associate with a particular AP

[Alex Hacker]

Hi guys!
There is a two APs with the same ESSID on the same frequency. The first has a
signal level 100000 times lower then the second one. The AR9285 equipped PC
is more sensetive or has better antenna then others. It tries to connect
to very weak neighboring AP with probably invalid key. I don't know why
wpa_suppicant prefer incorrect AP but IMO it is not a driver issue.
Simpliest solution is to select different ESSID for your AP.
Regards,
Alex.

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Hi Alex,

Le lundi 24 oct. 2011 ? 08:27:16 (+0200 CEST), Alex Hacker a ?crit?:
> Hi guys!
> There is a two APs with the same ESSID on the same frequency. The first has a
> signal level 100000 times lower then the second one. The AR9285 equipped PC
> is more sensetive or has better antenna then others. It tries to connect
> to very weak neighboring AP with probably invalid key. I don't know why
> wpa_suppicant prefer incorrect AP but IMO it is not a driver issue.
> Simpliest solution is to select different ESSID for your AP.

No, I can confirm there is only one AP with the same ESSID.

However, the same AP has several ESSIDs on the same channel (some being
hidden), with the following MAC addresses. The first 4 cells are "mine":

% iwlist wlan0 scan | grep -e "Cell" -e "ESSID"
          Cell 01 - Address: C6:6E:27:A6:BE:A4
                    ESSID:"Kirya"
          Cell 02 - Address: C6:6E:27:A6:BE:A5
                    ESSID:""
          Cell 03 - Address: C6:6E:27:A6:BE:A7
                    ESSID:""
          Cell 04 - Address: C6:6E:27:A6:BE:A6
                    ESSID:"FreeWifi"
          Cell 05 - Address: 5C:33:8E:EE:80:F9
                    ESSID:"Livebox-7bd6"
          Cell 06 - Address: 6A:17:33:4D:E4:4D
                    ESSID:"SFR WiFi Public"
          Cell 07 - Address: 00:24:D4:62:A0:BD
                    ESSID:"FreeWifi"
          Cell 08 - Address: 00:24:D4:60:F5:09
                    ESSID:"FreeWifi"
          Cell 09 - Address: 00:19:7D:78:69:89
                    ESSID:"Livebox-A380"
          Cell 10 - Address: 00:0A:78:9F:11:16
                    ESSID:"OLITEC"
          Cell 11 - Address: 00:24:D4:60:F5:08
                    ESSID:"freebox_DIZIEN"
          Cell 12 - Address: 00:17:33:4D:E4:4C
                    ESSID:"NEUF_E448"
          Cell 13 - Address: 6A:17:33:D0:86:C5
                    ESSID:"SFR WiFi Public"
          Cell 14 - Address: 00:16:41:8C:72:BB
                    ESSID:"Livebox-f3f9"
          Cell 15 - Address: 00:24:D4:62:A0:BC
                    ESSID:"Freebox-A6F7EB"

To answer another question about the MAC address of the ethernet interface,
it is 00:24:d4:bd:43:43

Neighbours with the same internet provide (and most probably the same
AP/router) all have APs with the MAC address beginning with 00:24:D4.
This prefix belongs to Freebox SA, my internet provider. The prefix C6:6E:27
is indeed unknown. Googling it leads to my posts on other mailing-lists, but
also some other posts amongst which one bug report [0] the logs of which
shows the exact same MAC address as my AP?
It also describes a similar problem than I have with a different module
(rt61pci) - the problem seems to have been fixed by installing the
firmware needed by the wlan card.

Cheers,
Julien

[0] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=563190

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Jouni Malinen]

On Mon, Oct 24, 2011 at 08:59:32PM +0200, Julien Valroff wrote:
> However, the same AP has several ESSIDs on the same channel (some being
> hidden), with the following MAC addresses. The first 4 cells are "mine":
> 
> % iwlist wlan0 scan | grep -e "Cell" -e "ESSID"
>           Cell 01 - Address: C6:6E:27:A6:BE:A4
>                     ESSID:"Kirya"
>           Cell 02 - Address: C6:6E:27:A6:BE:A5
>                     ESSID:""
>           Cell 03 - Address: C6:6E:27:A6:BE:A7
>                     ESSID:""
>           Cell 04 - Address: C6:6E:27:A6:BE:A6
>                     ESSID:"FreeWifi"

These seem to be locally administered addresses and you seemed to report
different addresses in previous messages... That's quite interesting AP
behavior..

> This prefix belongs to Freebox SA, my internet provider. The prefix C6:6E:27
> is indeed unknown. Googling it leads to my posts on other mailing-lists, but
> also some other posts amongst which one bug report [0] the logs of which
> shows the exact same MAC address as my AP?

Lovely.. While that is obviously possible with locally administered
addresses, getting two of those APs within range of each other would
result in really lovely mess for the stations.

It would be very useful to get a sniffer capture log from a separate
radio to see what exactly is happening. I think you mentioned that you
have number of other devices that could potentially be used for this,
but there was some issues with Wireshark. What radios/drivers do the
other devices have? Do they support IEEE 802.11n/HT? (If not, can you
temporarily disable that in the AP for the test?)

It looked like association with the AP went through, but the following
4-way handshake failed for some reason. Based on the information you've
sent so far, it looks like the station did not receive (or the AP did
not transmit) EAPOL-Key frames. The odd part was that you did not seem
to have issue if you connected this station as the first device to the
AP..

You do not need to use Wireshark for the sniffer setup. Assuming the
driver supports monitor mode, you can do this from the command line as
root:

ifconfig wlan0 down
iw wlan0 set type monitor
iw wlan0 set channel 6 HT20
ifconfig wlan0 up
tcpdump -i wlan0 -s 2500 -w /tmp/capture.dump

(with the channel obviously set to the one that the AP is using)

Then try to connect the AR9285-based station and once that fails, kill
tcpdump and send the capture log. If you can easily run this twice,
i.e., first with the success case where the station connects as the
first device and then with the failure case, this would be even more
valuable since it would make it easier to compare the message sequences.

-- 
Jouni Malinen                                            PGP id EFC895FA

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le lundi 24 oct. 2011 ? 21:55:01 (+0200 CEST), Jouni Malinen a ?crit?:
[...]
> You do not need to use Wireshark for the sniffer setup. Assuming the
> driver supports monitor mode, you can do this from the command line as
> root:
> 
> ifconfig wlan0 down
> iw wlan0 set type monitor
> iw wlan0 set channel 6 HT20
> ifconfig wlan0 up
> tcpdump -i wlan0 -s 2500 -w /tmp/capture.dump

I cannot get anything to be captured - I have tried from 2 laptops using
different drivers with no success. I even manage to stale the whole laptop
after bringing the interface up the second time when in monitoring mode.

Should the interface be associated when capturing? If so, how is this
achievable? I cannot set the essid then.

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le jeudi 27 oct. 2011 ? 20:36:06 (+0200 CEST), Julien Valroff a ?crit?:
> Le lundi 24 oct. 2011 ? 21:55:01 (+0200 CEST), Jouni Malinen a ?crit?:
> [...]
> > You do not need to use Wireshark for the sniffer setup. Assuming the
> > driver supports monitor mode, you can do this from the command line as
> > root:
> > 
> > ifconfig wlan0 down
> > iw wlan0 set type monitor
> > iw wlan0 set channel 6 HT20
> > ifconfig wlan0 up
> > tcpdump -i wlan0 -s 2500 -w /tmp/capture.dump
> 
> I cannot get anything to be captured - I have tried from 2 laptops using
> different drivers with no success. I even manage to stale the whole laptop
> after bringing the interface up the second time when in monitoring mode.

OK, I have finally managed to capture something. I had to add a monitor
interface: iw dev wlan0 interface add mon0 type monitor 

I attach both failure and success dumps, in the hope someone can detect
something.

I still had not time to set up FreeBSD, PC-BSD installer cannot start X on
my laptop (I have tried both stable and RC releases). Hopefully I'll be able
to do this during the next week-end.

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1
-------------- next part --------------
A non-text attachment was scrubbed...
Name: failed.dump.bz2
Type: application/octet-stream
Size: 19852 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111027/eab5dbf4/attachment-0002.obj 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: success.dump.bz2
Type: application/octet-stream
Size: 6964 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20111027/eab5dbf4/attachment-0003.obj 

[ath9k-devel] Can't associate with a particular AP

[Jouni Malinen]

On Thu, Oct 27, 2011 at 09:13:53PM +0200, Julien Valroff wrote:
> Le jeudi 27 oct. 2011 ? 20:36:06 (+0200 CEST), Julien Valroff a ?crit?:
> > Le lundi 24 oct. 2011 ? 21:55:01 (+0200 CEST), Jouni Malinen a ?crit?:
> > > ifconfig wlan0 down
> > > iw wlan0 set type monitor
> > > iw wlan0 set channel 6 HT20
> > > ifconfig wlan0 up
> > > tcpdump -i wlan0 -s 2500 -w /tmp/capture.dump
> > 
> > I cannot get anything to be captured - I have tried from 2 laptops using
> > different drivers with no success. I even manage to stale the whole laptop
> > after bringing the interface up the second time when in monitoring mode.

Does any of those commands fail? Which drivers did you use in these
tests?

> OK, I have finally managed to capture something. I had to add a monitor
> interface: iw dev wlan0 interface add mon0 type monitor 
> 
> I attach both failure and success dumps, in the hope someone can detect
> something.

Unfortunately, these capture files do not seem to include all frames.
They include mainly information that the sniffer device would have
received normally if it were in station mode. As such, I don't think it
was set to proper monitor mode. Which driver was used in this case?

-- 
Jouni Malinen                                            PGP id EFC895FA

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> Should the interface be associated when capturing?

No, rather the opposite. Ideally it should be passive in the wireless
network that it is monitoring. I've had good success with monitoring
mode and various Intel wifi chipsets.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le jeudi 27 oct. 2011 ? 22:27:58 (+0200 CEST), Jouni Malinen a ?crit?:
> On Thu, Oct 27, 2011 at 09:13:53PM +0200, Julien Valroff wrote:
> > Le jeudi 27 oct. 2011 ? 20:36:06 (+0200 CEST), Julien Valroff a ?crit?:
> > > Le lundi 24 oct. 2011 ? 21:55:01 (+0200 CEST), Jouni Malinen a ?crit?:
> > > > ifconfig wlan0 down
> > > > iw wlan0 set type monitor
> > > > iw wlan0 set channel 6 HT20
> > > > ifconfig wlan0 up
> > > > tcpdump -i wlan0 -s 2500 -w /tmp/capture.dump
> > > 
> > > I cannot get anything to be captured - I have tried from 2 laptops using
> > > different drivers with no success. I even manage to stale the whole laptop
> > > after bringing the interface up the second time when in monitoring mode.
> 
> Does any of those commands fail?

No, everything seems ok.

> Which drivers did you use in these tests?

brcmsmac and b43

> > OK, I have finally managed to capture something. I had to add a monitor
> > interface: iw dev wlan0 interface add mon0 type monitor 
> > 
> > I attach both failure and success dumps, in the hope someone can detect
> > something.
> 
> Unfortunately, these capture files do not seem to include all frames.
> They include mainly information that the sniffer device would have
> received normally if it were in station mode. As such, I don't think it
> was set to proper monitor mode.

mon0 which was captured was in monitoring mode, but not wlan0, which
might explain this problem. But when setting wlan0 in monitoring mode, I
cannot get anything captured.

> Which driver was used in this case?

brcmsmac

Thanks again for your time.

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Jouni Malinen]

On Fri, Oct 28, 2011 at 06:26:53AM +0200, Julien Valroff wrote:
> Le jeudi 27 oct. 2011 ? 22:27:58 (+0200 CEST), Jouni Malinen a ?crit?:
> > Which drivers did you use in these tests?
> 
> brcmsmac and b43

brcmsmac does not support monitor mode. b43 is claimed to support that,
but I've never tried. Though, if you are using a 802.11g card, it won't
be able to show IEEE 802.11n frames. If you happen to have any other
wireless cards, they could be more suitable for this purpose.

> mon0 which was captured was in monitoring mode, but not wlan0, which
> might explain this problem. But when setting wlan0 in monitoring mode, I
> cannot get anything captured.
> 
> > Which driver was used in this case?
> 
> brcmsmac

You won't get the this working with that driver until monitor mode
support has been added to it..

-- 
Jouni Malinen                                            PGP id EFC895FA

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le vendredi 28 oct. 2011 ? 10:35:19 (+0200 CEST), Jouni Malinen a ?crit?:
> On Fri, Oct 28, 2011 at 06:26:53AM +0200, Julien Valroff wrote:
> > Le jeudi 27 oct. 2011 ? 22:27:58 (+0200 CEST), Jouni Malinen a ?crit?:
> > > Which drivers did you use in these tests?
> > 
> > brcmsmac and b43
> 
> brcmsmac does not support monitor mode. b43 is claimed to support that,
> but I've never tried. Though, if you are using a 802.11g card, it won't
> be able to show IEEE 802.11n frames. If you happen to have any other
> wireless cards, they could be more suitable for this purpose.

Unfortunately, I do not have anything on hand. I'll try and check if I can
borrow another laptop and use a live cd but I'm not sure of the wlan card of
this laptop.

> > mon0 which was captured was in monitoring mode, but not wlan0, which
> > might explain this problem. But when setting wlan0 in monitoring mode, I
> > cannot get anything captured.
> > 
> > > Which driver was used in this case?
> > 
> > brcmsmac
> 
> You won't get the this working with that driver until monitor mode
> support has been added to it..

I have just tried again with b43, and got nothing captured. I have also
tried adding a monitoring interface, but got the same things as captured
yesterday using brcmsmac.

I'll try FreeBSD during the week-end.

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> > If you happen to have any other wireless cards, they could be
> > more suitable for this purpose.
> 
> Unfortunately, I do not have anything on hand.

Maybe someone nearby has a CardBus adapter that you can borrow and
use for monitoring, if not a complete laptop.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le lundi 24 oct. 2011 ? 20:59:32 (+0200 CEST), Julien Valroff a ?crit?:
> Hi Alex,
> 
> Le lundi 24 oct. 2011 ? 08:27:16 (+0200 CEST), Alex Hacker a ?crit?:
> > Hi guys!
> > There is a two APs with the same ESSID on the same frequency. The first has a
> > signal level 100000 times lower then the second one. The AR9285 equipped PC
> > is more sensetive or has better antenna then others. It tries to connect
> > to very weak neighboring AP with probably invalid key. I don't know why
> > wpa_suppicant prefer incorrect AP but IMO it is not a driver issue.
> > Simpliest solution is to select different ESSID for your AP.
> 
> No, I can confirm there is only one AP with the same ESSID.
> 
> However, the same AP has several ESSIDs on the same channel (some being
> hidden), with the following MAC addresses. The first 4 cells are "mine":
> 
> % iwlist wlan0 scan | grep -e "Cell" -e "ESSID"
>           Cell 01 - Address: C6:6E:27:A6:BE:A4
>                     ESSID:"Kirya"
>           Cell 02 - Address: C6:6E:27:A6:BE:A5
>                     ESSID:""
>           Cell 03 - Address: C6:6E:27:A6:BE:A7
>                     ESSID:""
>           Cell 04 - Address: C6:6E:27:A6:BE:A6
>                     ESSID:"FreeWifi"
>           Cell 05 - Address: 5C:33:8E:EE:80:F9
>                     ESSID:"Livebox-7bd6"
>           Cell 06 - Address: 6A:17:33:4D:E4:4D
>                     ESSID:"SFR WiFi Public"
>           Cell 07 - Address: 00:24:D4:62:A0:BD
>                     ESSID:"FreeWifi"
>           Cell 08 - Address: 00:24:D4:60:F5:09
>                     ESSID:"FreeWifi"
>           Cell 09 - Address: 00:19:7D:78:69:89
>                     ESSID:"Livebox-A380"
>           Cell 10 - Address: 00:0A:78:9F:11:16
>                     ESSID:"OLITEC"
>           Cell 11 - Address: 00:24:D4:60:F5:08
>                     ESSID:"freebox_DIZIEN"
>           Cell 12 - Address: 00:17:33:4D:E4:4C
>                     ESSID:"NEUF_E448"
>           Cell 13 - Address: 6A:17:33:D0:86:C5
>                     ESSID:"SFR WiFi Public"
>           Cell 14 - Address: 00:16:41:8C:72:BB
>                     ESSID:"Livebox-f3f9"
>           Cell 15 - Address: 00:24:D4:62:A0:BC
>                     ESSID:"Freebox-A6F7EB"

          Cell 01 - Address: 82:71:98:B6:C3:76
                    ESSID:"FreeWifi"
          Cell 02 - Address: 82:71:98:B6:C3:74
                    ESSID:"Kirya"
          Cell 03 - Address: 82:71:98:B6:C3:75
                    ESSID:""
          Cell 04 - Address: 82:71:98:B6:C3:77
                    ESSID:""
          Cell 05 - Address: 00:24:D4:62:A0:BD
                    ESSID:"FreeWifi"
          Cell 06 - Address: 5C:33:8E:EE:80:F9
                    ESSID:"Livebox-7bd6"
          Cell 07 - Address: 00:19:70:5D:E0:03
                    ESSID:"Livebox-2418"
          Cell 08 - Address: 00:24:D4:60:F5:08
                    ESSID:"freebox_DIZIEN"
          Cell 09 - Address: 00:16:41:8C:72:BB
                    ESSID:"Livebox-f3f9"
          Cell 10 - Address: 00:19:7D:78:69:89
                    ESSID:"Livebox-A380"
          Cell 11 - Address: 00:18:E7:3C:B8:92
                    ESSID:"Livebox-0EE8"

This doesn't cause any issue with the other drivers I use, but couldn't it
be the cause of the issues with ath9k?

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Adrian Chadd]

I bet it's not an ath9k problem. I bet you're just finding the really
high receive sensitivity of the atheros NICs. :)

Ie, ath9k is not choosing which AP to associate with. I bet that
wpa_supplicant / 802.11 stack is not blacklisting that far-away AP:

* because the S/N is too low;
* repeated, failed attempts at association.

I know FreeBSD's net80211 stack has some code in it which rejects
things if the two above conditions are met. I think the Linux
wpa_supplicant code (which I bet is what's making decisions as to the
BSSID's to be associating against, but I'm not _that_ familiar with
the workings of the Linux stack just yet) needs to be tinkered with a
bit more.

So, bounce this problem and what we've discovered thus far to the
linux-wireless mailing list and let them at it. :)



Adrian

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Adrian Chadd wrote:
> Ie, ath9k is not choosing which AP to associate with.

Well, it can..


> I think the Linux wpa_supplicant code (which I bet is what's making
> decisions as to the BSSID's to be associating against,

I can also associate without wpa_supplicant, if the AP does not
require WPA.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Julien Valroff]

Le samedi 29 oct. 2011 ? 18:27:28 (+0200 CEST), Peter Stuge a ?crit?:
> Adrian Chadd wrote:
> > Ie, ath9k is not choosing which AP to associate with.
> 
> Well, it can..

How can you force that?
I've tried doing this in NetworkManager with no success (ie. same problems
as without).

> > I think the Linux wpa_supplicant code (which I bet is what's making
> > decisions as to the BSSID's to be associating against,
> 
> I can also associate without wpa_supplicant, if the AP does not
> require WPA.

It also fails to associate with my AP when no security is used (as
explained, I have several SSID's on the same AP, one of them being used as a
hotspot and gives the same issues as my private wlan).

Cheers,
Julien

-- 
  .''`.   Julien Valroff ~ <julien@kirya.net> ~ <julien@debian.org>    
 : :'  :  Debian Developer & Free software contributor
 `. `'`   http://www.kirya.net/
   `-     4096R/ E1D8 5796 8214 4687 E416  948C 859F EF67 258E 26B1

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Julien Valroff wrote:
> > > Ie, ath9k is not choosing which AP to associate with.
> > 
> > Well, it can..
> 
> How can you force that?

Force what, exactly?


> I've tried doing this in NetworkManager with no success (ie. same
> problems as without).

I would strongly avoid NetworkManager and wpa_supplicant while
debugging. Because they tie in to the wifi stack and poke at things
they think are good, I find them and their kin mostly a neccessary
evil.

Use iw and/or iwconfig.

http://linuxwireless.org/en/users/Documentation/iw#Establishing_a_basic_connection


> > > I think the Linux wpa_supplicant code (which I bet is what's making
> > > decisions as to the BSSID's to be associating against,
> > 
> > I can also associate without wpa_supplicant, if the AP does not
> > require WPA.
> 
> It also fails to associate with my AP when no security is used (as
> explained, I have several SSID's on the same AP, one of them being
> used as a hotspot and gives the same issues as my private wlan).

Do not bother with anything that has encryption disabled unless
neccessary (not in this case) since it adds a significant amount
of unknowns.

It's actually really good imo that you can see the problem also with
the AP without security.


//Peter

[ath9k-devel] Can't associate with a particular AP

[Adrian Chadd]

On 30 October 2011 03:27, Peter Stuge <peter@stuge.se> wrote:
> Adrian Chadd wrote:
>> Ie, ath9k is not choosing which AP to associate with.
>
> Well, it can..

How? All ath9k is doing is providing the raw frames to mac80211. The
decision as to what to associate to is done further up in the stack.
(The same is done in FreeBSD - the ath(4) driver only provides the raw
frames to the net80211 layer; that and wpa_supplicant determine what
to associate to.)



Adrian

[ath9k-devel] Can't associate with a particular AP

[Peter Stuge]

Adrian Chadd wrote:
> >> Ie, ath9k is not choosing which AP to associate with.
> >
> > Well, it can..
> 
> How? All ath9k is doing is providing the raw frames to mac80211.

By "it" I of course mean "the kernel" :p


//Peter