[ath9k-devel] trouble with DWA-552 and ath9k

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

I recently purchased a DLink DWA-552 that I would like to use with 
hostapd as an access point. The ath9k driver claims the chipset is:

Atheros AR9280 MAC/BB Rev:2 AR5133 RF Rev:d0

I have tried both the ath9k that comes with my distribution (Fedora 11) 
in the 2.6.30.9-102.fc11.i686.PAE kernel and also downloading, building 
and installing the latest compat-wireless-2.6.tar.bz2 with date 
2009-12-11 and rebooting.

In both cases, I am unable to find any access points when doing an 
iwlist wlan0 scan while other 802.11 devices in my home find several 
access points from nearby neighbors. I have also tried both hostapd 
0.6.9 and 0.7.0. When I try to run hostapd to put the card into AP mode, 
I get the error:

nl80211: Failed to set channel (freq=2412): -22 (Invalid argument)

I don't know if it is supposed to, but the LED on the card itself never 
comes on.

Information/Logs:

iw phy0 info
Wiphy phy0
        Band 1:
                HT capabilities: 0x104e
                        * 20/40 MHz operation
                        * SM PS disabled
                        * 40 MHz short GI
                        * max A-MSDU len 3839
                        * DSSS/CCK 40 MHz
                HT A-MPDU factor: 0x0003 (65535 bytes)
                HT A-MPDU density: 0x0006 (8 usec)
                HT MCS set: ff ff 00 00 00 00 00 00 00 00 00 00 01 00 00 00
                HT TX/RX MCS rate indexes supported:
                        MCS index 0
                        MCS index 1
                        MCS index 2
                        MCS index 3
                        MCS index 4
                        MCS index 5
                        MCS index 6
                        MCS index 7
                        MCS index 8
                        MCS index 9
                        MCS index 10
                        MCS index 11
                        MCS index 12
                        MCS index 13
                        MCS index 14
                        MCS index 15
                Frequencies:
                        * 2412 MHz [1] (27.0 dBm)
                        * 2417 MHz [2] (27.0 dBm)
                        * 2422 MHz [3] (27.0 dBm)
                        * 2427 MHz [4] (27.0 dBm)
                        * 2432 MHz [5] (27.0 dBm)
                        * 2437 MHz [6] (27.0 dBm)
                        * 2442 MHz [7] (27.0 dBm)
                        * 2447 MHz [8] (27.0 dBm)
                        * 2452 MHz [9] (27.0 dBm)
                        * 2457 MHz [10] (27.0 dBm)
                        * 2462 MHz [11] (27.0 dBm)
                        * 2467 MHz [12] (disabled)
                        * 2472 MHz [13] (disabled)
                        * 2484 MHz [14] (disabled)
                Bitrates (non-HT):
                        * 1.0 Mbps
                        * 2.0 Mbps (short preamble supported)
                        * 5.5 Mbps (short preamble supported)
                        * 11.0 Mbps (short preamble supported)
                        * 6.0 Mbps
                        * 9.0 Mbps
                        * 12.0 Mbps
                        * 18.0 Mbps
                        * 24.0 Mbps
                        * 36.0 Mbps
                        * 48.0 Mbps
                        * 54.0 Mbps
        max # scan SSIDs: 4
        Supported interface modes:
                 * IBSS
                 * managed
                 * AP
                 * AP/VLAN
                 * monitor
                 * mesh point

hostapd -dd test.conf
Configuration file: test.conf
nl80211: Add own interface ifindex 4
nl80211: Add own interface ifindex 63
BSS count 1, BSSID mask 00:00:00:00:00:00 (0 bits)
nl80211: Added 802.11b mode based on 802.11g information
Allowed channel: mode=1 chan=1 freq=2412 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=2 freq=2417 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=3 freq=2422 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=4 freq=2427 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=5 freq=2432 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=6 freq=2437 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=7 freq=2442 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=8 freq=2447 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=9 freq=2452 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=10 freq=2457 MHz max_tx_power=27 dBm
Allowed channel: mode=1 chan=11 freq=2462 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=1 freq=2412 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=2 freq=2417 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=3 freq=2422 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=4 freq=2427 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=5 freq=2432 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=6 freq=2437 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=7 freq=2442 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=8 freq=2447 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=9 freq=2452 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=10 freq=2457 MHz max_tx_power=27 dBm
Allowed channel: mode=0 chan=11 freq=2462 MHz max_tx_power=27 dBm
RATE[0] rate=10 flags=0x1
RATE[1] rate=20 flags=0x1
RATE[2] rate=55 flags=0x1
RATE[3] rate=110 flags=0x1
RATE[4] rate=60 flags=0x0
RATE[5] rate=90 flags=0x0
RATE[6] rate=120 flags=0x0
RATE[7] rate=180 flags=0x0
RATE[8] rate=240 flags=0x0
RATE[9] rate=360 flags=0x0
RATE[10] rate=480 flags=0x0
RATE[11] rate=540 flags=0x0
Completing interface initialization
Mode: IEEE 802.11g  Channel: 1  Frequency: 2412 MHz
nl80211: Failed to set channel (freq=2412): -22 (Invalid argument)
Could not set channel for kernel driver
wlan0: Unable to setup interface.
Flushing old station entries
Deauthenticate all stations
nl80211: Operstate: linkmode=0, operstate=6

test.conf:
interface=wlan0
driver=nl80211

hw_mode=g
channel=1
ssid=test

wpa=2
wpa_key_mgmt=WPA-PSK
wpa_pairwise=CCMP
wpa_passphrase=12345678

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> I recently purchased a DLink DWA-552 that I would like to use with
> hostapd as an access point. The ath9k driver claims the chipset is:
> 
> Atheros AR9280 MAC/BB Rev:2 AR5133 RF Rev:d0
> 
> I have tried both the ath9k that comes with my distribution (Fedora 11)
> in the 2.6.30.9-102.fc11.i686.PAE kernel and also downloading, building
> and installing the latest compat-wireless-2.6.tar.bz2 with date
> 2009-12-11 and rebooting.

Can you build compat-wireless with debug enabled ?
In config.mk, set CONFIG_ATH9K_DEBUG=y, rebuild and reload the driver
with the module parameter "debug=0x600" and post the kernel log.

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Hi. In config.mk I find no option named CONFIG_ATH9K_DEBUG, but there is 
one named CONFIG_ATH9K_DEBUGFS. Also, as I tried to include in a 
followup to my original post, the output from lspci -vnn is:

00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X 
Wireless Network Adapter [168c:0029] (rev 01)
     Subsystem: D-Link System Inc Device [1186:3a7a]
     Flags: 66MHz, medium devsel, IRQ 5
     [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
     Capabilities: [44] Power Management version 2
     Kernel driver in use: ath9k
     Kernel modules: ath9k

The 168c:0029 seems pretty standard for an Atheros wireless N adapter, 
but I wonder if the 1186:3a7a is a new revision of this card and is 
perhaps not 100% supported yet?

Brian

On 12/17/2009 10:12 PM, Sujith wrote:
> Brian Walker wrote:
>    
>> I recently purchased a DLink DWA-552 that I would like to use with
>> hostapd as an access point. The ath9k driver claims the chipset is:
>>
>> Atheros AR9280 MAC/BB Rev:2 AR5133 RF Rev:d0
>>
>> I have tried both the ath9k that comes with my distribution (Fedora 11)
>> in the 2.6.30.9-102.fc11.i686.PAE kernel and also downloading, building
>> and installing the latest compat-wireless-2.6.tar.bz2 with date
>> 2009-12-11 and rebooting.
>>      
> Can you build compat-wireless with debug enabled ?
> In config.mk, set CONFIG_ATH9K_DEBUG=y, rebuild and reload the driver
> with the module parameter "debug=0x600" and post the kernel log.
>
> Sujith
>
>
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> Hi. In config.mk I find no option named CONFIG_ATH9K_DEBUG, but there is
> one named CONFIG_ATH9K_DEBUGFS. Also, as I tried to include in a
> followup to my original post, the output from lspci -vnn is:

Indeed, the option is missing.
Adding the line "CONFIG_ATH_DEBUG=y" and "CONFIG_ATH9K_DEBUGFS=y" to config.mk
and recompiling should do it.

> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X
> Wireless Network Adapter [168c:0029] (rev 01)
>      Subsystem: D-Link System Inc Device [1186:3a7a]
>      Flags: 66MHz, medium devsel, IRQ 5
>      [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>      Capabilities: [44] Power Management version 2
>      Kernel driver in use: ath9k
>      Kernel modules: ath9k
> 
> The 168c:0029 seems pretty standard for an Atheros wireless N adapter,
> but I wonder if the 1186:3a7a is a new revision of this card and is
> perhaps not 100% supported yet?

That can be determined by looking at the kernel log on loading the driver
with 'debug=0x600'.

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Thanks for that information! I'm recompiling ath9k with 
CONFIG_ATH9K_DEBUG=y and CONFIG_ATH9K_DEBUGFS=y. After that I will 
reboot and load ath9k with debug=0x600. Looking at 
http://linuxwireless.org/en/users/Drivers/ath9k/debug for information on 
debugging the ath9k driver, is there anything of particular interest in 
/sys/kernel/debug/ath9k that I should post?

Brian

On 12/17/2009 11:44 PM, Sujith wrote:
> Brian Walker wrote:
>    
>> Hi. In config.mk I find no option named CONFIG_ATH9K_DEBUG, but there is
>> one named CONFIG_ATH9K_DEBUGFS. Also, as I tried to include in a
>> followup to my original post, the output from lspci -vnn is:
>>      
> Indeed, the option is missing.
> Adding the line "CONFIG_ATH_DEBUG=y" and "CONFIG_ATH9K_DEBUGFS=y" to config.mk
> and recompiling should do it.
>
>    
>> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X
>> Wireless Network Adapter [168c:0029] (rev 01)
>>       Subsystem: D-Link System Inc Device [1186:3a7a]
>>       Flags: 66MHz, medium devsel, IRQ 5
>>       [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>>       Capabilities: [44] Power Management version 2
>>       Kernel driver in use: ath9k
>>       Kernel modules: ath9k
>>
>> The 168c:0029 seems pretty standard for an Atheros wireless N adapter,
>> but I wonder if the 1186:3a7a is a new revision of this card and is
>> perhaps not 100% supported yet?
>>      
> That can be determined by looking at the kernel log on loading the driver
> with 'debug=0x600'.
>
> Sujith
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

I've recompiled compat-wireless-2009-12-11 with "CONFIG_ATH_DEBUG=y" and 
"CONFIG_ATH9K_DEBUGFS=y" in config.mk (and did make and then make 
install), added  "options ath9k debug=0x600" to the file 
/etc/modprobe.d/local.conf (I am running Fedora 11) and rebooted.

After rebooting, I did "mount -t debugfs none /sys/kernel/debug/". 
Indeed, there is now a /sys/kernel/debug/ath9k directory. Am I looking 
for anything in particular? The directory /sys/kernel/debug/ath9k/phy0/ 
has the files:

dma  interrupt  rcstat  wiphy  xmit

I'm attaching the output of dmesg right after booting.

Couple of other observations... If I rmmod ath9k and then try to 
modprobe/insmod it again, I get an error that ath9k "failed to 
initialize device" and "PCI INT A disabled".  Sometimes after rebooting 
and trying to use wlan0 the entire system just locks up. The most recent 
time this happened, I had just booted up and issued the command 
"ifconfig wlan0 up". I was going to next issue the command "iwlist wlan0 
scan", but the system had already locked up. The only way to recover is 
to hit the reset button.

Brian

On 12/17/2009 11:44 PM, Sujith wrote:
> Brian Walker wrote:
>    
>> Hi. In config.mk I find no option named CONFIG_ATH9K_DEBUG, but there is
>> one named CONFIG_ATH9K_DEBUGFS. Also, as I tried to include in a
>> followup to my original post, the output from lspci -vnn is:
>>      
> Indeed, the option is missing.
> Adding the line "CONFIG_ATH_DEBUG=y" and "CONFIG_ATH9K_DEBUGFS=y" to config.mk
> and recompiling should do it.
>
>    
>> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X
>> Wireless Network Adapter [168c:0029] (rev 01)
>>       Subsystem: D-Link System Inc Device [1186:3a7a]
>>       Flags: 66MHz, medium devsel, IRQ 5
>>       [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>>       Capabilities: [44] Power Management version 2
>>       Kernel driver in use: ath9k
>>       Kernel modules: ath9k
>>
>> The 168c:0029 seems pretty standard for an Atheros wireless N adapter,
>> but I wonder if the 1186:3a7a is a new revision of this card and is
>> perhaps not 100% supported yet?
>>      
> That can be determined by looking at the kernel log on loading the driver
> with 'debug=0x600'.
>
> Sujith
>    
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: dmesg.txt
Url: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20091218/c6f06892/attachment-0001.txt 

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> I've recompiled compat-wireless-2009-12-11 with "CONFIG_ATH_DEBUG=y" and
> "CONFIG_ATH9K_DEBUGFS=y" in config.mk (and did make and then make
> install), added  "options ath9k debug=0x600" to the file
> /etc/modprobe.d/local.conf (I am running Fedora 11) and rebooted.
> 
> After rebooting, I did "mount -t debugfs none /sys/kernel/debug/".
> Indeed, there is now a /sys/kernel/debug/ath9k directory. Am I looking
> for anything in particular? The directory /sys/kernel/debug/ath9k/phy0/
> has the files:
> 
> dma  interrupt  rcstat  wiphy  xmit
> 
> I'm attaching the output of dmesg right after booting.
> 
> Couple of other observations... If I rmmod ath9k and then try to
> modprobe/insmod it again, I get an error that ath9k "failed to
> initialize device" and "PCI INT A disabled".  Sometimes after rebooting
> and trying to use wlan0 the entire system just locks up. The most recent
> time this happened, I had just booted up and issued the command
> "ifconfig wlan0 up". I was going to next issue the command "iwlist wlan0
> scan", but the system had already locked up. The only way to recover is
> to hit the reset button.

Does the kernel lockup every time the interface is brought up ?
If not, can you post the kernel log on doing 'ifconfig wlan0 up' ?

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 12/22/2009 09:10 AM, Brian Walker wrote:
> cfg80211: World regulatory domain updated:
>     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
>     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
>     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
> ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 5
> PCI: setting IRQ 5 as level-triggered
> ath9k 0000:00:0a.0: PCI INT A -> Link[LNKC] -> GSI 5 (level, low) -> 
> IRQ 5
> ath: EEPROM regdomain: 0x10
> ath: EEPROM indicates we should expect a direct regpair map
> ath: Country alpha2 being used: CO
> ath: Regpair used: 0x10
> phy0: Selected rate control algorithm 'ath9k_rate_control'
> Registered led device: ath9k-phy0::radio
> Registered led device: ath9k-phy0::assoc
> Registered led device: ath9k-phy0::tx
> Registered led device: ath9k-phy0::rx
> phy0: Atheros AR9280 Rev:2 mem=0xfa4a0000, irq=5
> cfg80211: Calling CRDA for country: CO
> cfg80211: Regulatory domain changed to country: CO
>     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
>     (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
>     (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
>     (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
> ADDRCONF(NETDEV_UP): wlan0: link is not ready
>
> The system lockups seem to occur most often when the debugfs 
> filesystem is mounted. Also, curious why the driver is saying AR9280 
> when lspci says AR922X? In /etc/modprobe.d/local.conf I have now set:
>
> options ath9k debug=0xffffffff
>
> since I do not seem to be getting any kind of debugging messages, but 
> I still don't see much in the way of debugging messages in  the kernel 
> logs.
>
> Brian
>
> On 12/21/2009 11:58 PM, Sujith wrote:
>> Brian Walker wrote:
>>> I've recompiled compat-wireless-2009-12-11 with "CONFIG_ATH_DEBUG=y" 
>>> and
>>> "CONFIG_ATH9K_DEBUGFS=y" in config.mk (and did make and then make
>>> install), added  "options ath9k debug=0x600" to the file
>>> /etc/modprobe.d/local.conf (I am running Fedora 11) and rebooted.
>>>
>>> After rebooting, I did "mount -t debugfs none /sys/kernel/debug/".
>>> Indeed, there is now a /sys/kernel/debug/ath9k directory. Am I looking
>>> for anything in particular? The directory /sys/kernel/debug/ath9k/phy0/
>>> has the files:
>>>
>>> dma  interrupt  rcstat  wiphy  xmit
>>>
>>> I'm attaching the output of dmesg right after booting.
>>>
>>> Couple of other observations... If I rmmod ath9k and then try to
>>> modprobe/insmod it again, I get an error that ath9k "failed to
>>> initialize device" and "PCI INT A disabled".  Sometimes after rebooting
>>> and trying to use wlan0 the entire system just locks up. The most 
>>> recent
>>> time this happened, I had just booted up and issued the command
>>> "ifconfig wlan0 up". I was going to next issue the command "iwlist 
>>> wlan0
>>> scan", but the system had already locked up. The only way to recover is
>>> to hit the reset button.
>> Does the kernel lockup every time the interface is brought up ?
>> If not, can you post the kernel log on doing 'ifconfig wlan0 up' ?
>>
>> Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Can anyone else confirm that this card works (or definitely does not 
work) with the current (2009-12-11) ath9k driver?

00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X 
Wireless Network Adapter [168c:0029] (rev 01)
     Subsystem: D-Link System Inc Device [1186:3a7a]
     Flags: 66MHz, medium devsel, IRQ 5
     [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
     Capabilities: [44] Power Management version 2
     Kernel driver in use: ath9k
     Kernel modules: ath9k

Brian

On 12/17/2009 11:44 PM, Sujith wrote:
> Brian Walker wrote:
>    
>> Hi. In config.mk I find no option named CONFIG_ATH9K_DEBUG, but there is
>> one named CONFIG_ATH9K_DEBUGFS. Also, as I tried to include in a
>> followup to my original post, the output from lspci -vnn is:
>>      
> Indeed, the option is missing.
> Adding the line "CONFIG_ATH_DEBUG=y" and "CONFIG_ATH9K_DEBUGFS=y" to config.mk
> and recompiling should do it.
>
>    
>> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X
>> Wireless Network Adapter [168c:0029] (rev 01)
>>       Subsystem: D-Link System Inc Device [1186:3a7a]
>>       Flags: 66MHz, medium devsel, IRQ 5
>>       [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>>       Capabilities: [44] Power Management version 2
>>       Kernel driver in use: ath9k
>>       Kernel modules: ath9k
>>
>> The 168c:0029 seems pretty standard for an Atheros wireless N adapter,
>> but I wonder if the 1186:3a7a is a new revision of this card and is
>> perhaps not 100% supported yet?
>>      
> That can be determined by looking at the kernel log on loading the driver
> with 'debug=0x600'.
>
> Sujith
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Mon, 2009-12-28 at 14:22 -0500, Brian Walker wrote:
> Can anyone else confirm that this card works (or definitely does not 
> work) with the current (2009-12-11) ath9k driver?
> 
> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X 
> Wireless Network Adapter [168c:0029] (rev 01)
>      Subsystem: D-Link System Inc Device [1186:3a7a]
>      Flags: 66MHz, medium devsel, IRQ 5
>      [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>      Capabilities: [44] Power Management version 2
>      Kernel driver in use: ath9k
>      Kernel modules: ath9k

I can definitely confirm that DWA-552 is working for me.  It didn't work
on newer systems with Intel Quad Core CPU and Intel G35 chipset for
reasons unrelated to ath9k.  Either it's not seen by lspci, or the PCI
ID is corrupt.

However, DWA-552 is working with ath9k on older systems, such as an AMD
XP 2000 system with Asus V7B8X motherboard or PowerMac G3.  The access
points are seen in the scan results and WPA connection is established
quickly and reliably.

The system and subsystem IDs is different, but the card is marked as
D-Link DWA-552.

01:03.0 Network controller [0280]: Atheros Communications Inc. AR5008
Wireless Network Adapter [168c:0023] (rev 01)
        Subsystem: D-Link System Inc Device [1186:3a6d]
        Flags: bus master, 66MHz, medium devsel, latency 168, IRQ 24
        Memory at 80890000 (32-bit, non-prefetchable) [size=64K]
        Capabilities: [40] #80 [0000]
        Kernel driver in use: ath9k
        Kernel modules: ath9k

wlan0     IEEE 802.11bgn  Mode:Managed  Frequency:2.437 GHz  
          Access Point: Not-Associated   Tx-Power=27 dBm   
          Retry  long limit:7   RTS thr:off   Fragment thr:off
          Encryption key:off
          Power Management:on
          
# wpa_supplicant -c /etc/wpa_supplicant/mike1.conf -D wext -i wlan0
ioctl[SIOCGIWSCAN]: Resource temporarily unavailable
ioctl[SIOCGIWSCAN]: Resource temporarily unavailable
Trying to associate with 00:19:5b:56:fc:73 (SSID='mike1' freq=2437 MHz)
Associated with 00:19:5b:56:fc:73
WPA: Key negotiation completed with 00:19:5b:56:fc:73 [PTK=TKIP
GTK=TKIP]
CTRL-EVENT-CONNECTED - Connection to 00:19:5b:56:fc:73 completed (auth)
[id=0 id_str=]

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

My DWA-552 is in a P3 1ghz (very old) system with an AX63pro 
motherboard. I think the difference in subsystem id is perhaps the 
issue? Yours shows as AR5008 but mine shows as AR922X. Does anyone else 
have the exact same card as I do and it is working?

Brian

On 12/28/2009 03:49 PM, Pavel Roskin wrote:
> On Mon, 2009-12-28 at 14:22 -0500, Brian Walker wrote:
>    
>> Can anyone else confirm that this card works (or definitely does not
>> work) with the current (2009-12-11) ath9k driver?
>>
>> 00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X
>> Wireless Network Adapter [168c:0029] (rev 01)
>>       Subsystem: D-Link System Inc Device [1186:3a7a]
>>       Flags: 66MHz, medium devsel, IRQ 5
>>       [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
>>       Capabilities: [44] Power Management version 2
>>       Kernel driver in use: ath9k
>>       Kernel modules: ath9k
>>      
> I can definitely confirm that DWA-552 is working for me.  It didn't work
> on newer systems with Intel Quad Core CPU and Intel G35 chipset for
> reasons unrelated to ath9k.  Either it's not seen by lspci, or the PCI
> ID is corrupt.
>
> However, DWA-552 is working with ath9k on older systems, such as an AMD
> XP 2000 system with Asus V7B8X motherboard or PowerMac G3.  The access
> points are seen in the scan results and WPA connection is established
> quickly and reliably.
>
> The system and subsystem IDs is different, but the card is marked as
> D-Link DWA-552.
>
> 01:03.0 Network controller [0280]: Atheros Communications Inc. AR5008
> Wireless Network Adapter [168c:0023] (rev 01)
>          Subsystem: D-Link System Inc Device [1186:3a6d]
>          Flags: bus master, 66MHz, medium devsel, latency 168, IRQ 24
>          Memory at 80890000 (32-bit, non-prefetchable) [size=64K]
>          Capabilities: [40] #80 [0000]
>          Kernel driver in use: ath9k
>          Kernel modules: ath9k
>
> wlan0     IEEE 802.11bgn  Mode:Managed  Frequency:2.437 GHz
>            Access Point: Not-Associated   Tx-Power=27 dBm
>            Retry  long limit:7   RTS thr:off   Fragment thr:off
>            Encryption key:off
>            Power Management:on
>
> # wpa_supplicant -c /etc/wpa_supplicant/mike1.conf -D wext -i wlan0
> ioctl[SIOCGIWSCAN]: Resource temporarily unavailable
> ioctl[SIOCGIWSCAN]: Resource temporarily unavailable
> Trying to associate with 00:19:5b:56:fc:73 (SSID='mike1' freq=2437 MHz)
> Associated with 00:19:5b:56:fc:73
> WPA: Key negotiation completed with 00:19:5b:56:fc:73 [PTK=TKIP
> GTK=TKIP]
> CTRL-EVENT-CONNECTED - Connection to 00:19:5b:56:fc:73 completed (auth)
> [id=0 id_str=]
>
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Mon, 2009-12-28 at 16:06 -0500, Brian Walker wrote:
> My DWA-552 is in a P3 1ghz (very old) system with an AX63pro 
> motherboard. I think the difference in subsystem id is perhaps the 
> issue?

There is a difference in the system ID (23 vs 29), which is more
important.

>  Yours shows as AR5008 but mine shows as AR922X. Does anyone else 
> have the exact same card as I do and it is working?

I think you have an issue with the calibration.  I've seen it with
Ubiquity SR71A.  That's indeed AR922X, so it's not supported by MadWifi
at all.  With ath9k, the interface cannot be brought up the first time
(debugging should a calibration timeout).  It can be brought up the
second time (that is, the driver doesn't report an error), but there are
no scan results.

That card fails in FreeBSD 8.0 exactly in the same way, as if the code
was written from the same specification.  But it does work under Windows
XP with the Ubiquity driver (I had to patch the inf file, as it was
missing the subsystem ID for the card).

We could trace the calibration done by the Ubiquiti XP drivers under
ndiswrapper and compare it with the calibration done by ath9k to see the
difference.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Ok? I'd be glad to provide a trace, etc. if I were given instructions on 
what was desired and how to do so. As I stated when I started this 
chain, I get the following when trying to use hostapd:

nl80211: Failed to set channel (freq=2412): -22 (Invalid argument)

When trying to use the card in managed mode, it does indeed fail on the 
first "ifconfig wlan0 up" command with
"SIOCSIFFLAGS: Input/output error". The second time I issue the ifconfig 
command it succeeds. Running "iwlist wlan0 scan" still returns no 
results although other wireless devices pickup several access points 
from neighbors.

Be glad to do or provide anything that might get this card working.

Brian

On 12/28/2009 04:28 PM, Pavel Roskin wrote:
> On Mon, 2009-12-28 at 16:06 -0500, Brian Walker wrote:
>    
>> My DWA-552 is in a P3 1ghz (very old) system with an AX63pro
>> motherboard. I think the difference in subsystem id is perhaps the
>> issue?
>>      
> There is a difference in the system ID (23 vs 29), which is more
> important.
>
>    
>>   Yours shows as AR5008 but mine shows as AR922X. Does anyone else
>> have the exact same card as I do and it is working?
>>      
> I think you have an issue with the calibration.  I've seen it with
> Ubiquity SR71A.  That's indeed AR922X, so it's not supported by MadWifi
> at all.  With ath9k, the interface cannot be brought up the first time
> (debugging should a calibration timeout).  It can be brought up the
> second time (that is, the driver doesn't report an error), but there are
> no scan results.
>
> That card fails in FreeBSD 8.0 exactly in the same way, as if the code
> was written from the same specification.  But it does work under Windows
> XP with the Ubiquity driver (I had to patch the inf file, as it was
> missing the subsystem ID for the card).
>
> We could trace the calibration done by the Ubiquiti XP drivers under
> ndiswrapper and compare it with the calibration done by ath9k to see the
> difference.
>
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

Quoting Brian Walker <walkerbm@charter.net>:

> Ok? I'd be glad to provide a trace, etc. if I were given instructions
> on what was desired and how to do so.

It's easier for me to take the trace myself that to provide  
instructions.  Besides, the hardest part would be deciphering them and  
figuring out what is wrong in the code.

I see that your DWA-552 card is identified as
Atheros AR9280 MAC/BB Rev:2 AR5133 RF Rev:d0

and my Ubiquiti SR-71A is identified as
phy3: Atheros AR9280 Rev:2 mem=0xffffc90005940000, irq=19

I think the current wireless-testing doesn't print detailed chip info,  
as it knows that AR9280 and newer are single chips.

Anyway, we have Atheros AR9280 Rev 2 in both cases.  Also, the PCI ID  
is 168c:0029 for both cards that exhibit the problem.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Ok, so you're going to do testing/tracing and then perhaps submit a 
patch to get these cards to work based on findings?

Brian

On 12/29/2009 10:02 AM, Pavel Roskin wrote:
> Quoting Brian Walker <walkerbm@charter.net>:
>
>> Ok? I'd be glad to provide a trace, etc. if I were given instructions
>> on what was desired and how to do so.
>
> It's easier for me to take the trace myself that to provide 
> instructions.  Besides, the hardest part would be deciphering them and 
> figuring out what is wrong in the code.
>
> I see that your DWA-552 card is identified as
> Atheros AR9280 MAC/BB Rev:2 AR5133 RF Rev:d0
>
> and my Ubiquiti SR-71A is identified as
> phy3: Atheros AR9280 Rev:2 mem=0xffffc90005940000, irq=19
>
> I think the current wireless-testing doesn't print detailed chip info, 
> as it knows that AR9280 and newer are single chips.
>
> Anyway, we have Atheros AR9280 Rev 2 in both cases.  Also, the PCI ID 
> is 168c:0029 for both cards that exhibit the problem.
>

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

Quoting Brian Walker <walkerbm@charter.net>:

> Ok, so you're going to do testing/tracing and then perhaps submit a
> patch to get these cards to work based on findings?

Only if I find time for that.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Any update on getting this card working with the ath9k?

Brian

Pavel Roskin wrote:
> Quoting Brian Walker <walkerbm@charter.net>:
>
>> Ok, so you're going to do testing/tracing and then perhaps submit a
>> patch to get these cards to work based on findings?
>
> Only if I find time for that.
>

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> Any update on getting this card working with the ath9k?

The logs that you posted earlier didn't contain debug information.
Can you make sure that you load the driver with debug=0x601 and
post the dmesg output on doing 'ifconfig wlan0 up' ?

Please see: http://wireless.kernel.org/en/users/Drivers/ath9k/debug

That would show what exactly goes wrong when bringing up the interface.

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):

options ath9k debug=0xffffffff

In config.mk I have:

CONFIG_ATH9K_DEBUG=y
CONFIG_ATH9K_DEBUGFS=y

Is there something else I need to do to get the debugging information 
you're interested in?

Brian

On 01/07/2010 11:34 PM, Sujith wrote:
> Brian Walker wrote:
>    
>> Any update on getting this card working with the ath9k?
>>      
> The logs that you posted earlier didn't contain debug information.
> Can you make sure that you load the driver with debug=0x601 and
> post the dmesg output on doing 'ifconfig wlan0 up' ?
>
> Please see: http://wireless.kernel.org/en/users/Drivers/ath9k/debug
>
> That would show what exactly goes wrong when bringing up the interface.
>
> Sujith
>    

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Sat, 2010-01-09 at 10:14 -0500, Brian Walker wrote:
> This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):
> 
> options ath9k debug=0xffffffff
> 
> In config.mk I have:
> 
> CONFIG_ATH9K_DEBUG=y
> CONFIG_ATH9K_DEBUGFS=y
> 
> Is there something else I need to do to get the debugging information 
> you're interested in?
> 
> Brian
> 
> On 01/07/2010 11:34 PM, Sujith wrote:
> > Brian Walker wrote:
> >    
> >> Any update on getting this card working with the ath9k?
> >>      
> > The logs that you posted earlier didn't contain debug information.
> > Can you make sure that you load the driver with debug=0x601 and
> > post the dmesg output on doing 'ifconfig wlan0 up' ?

I think the quoted message is pretty clear.  Maybe it would be better
that you write under the quoted text so that you can see what you have
quoted?

debug=0xffffffff would likely produce too much information in the kernel
log, so please change it to 0x601 as you were asked.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/09/2010 10:38 AM, Pavel Roskin wrote:
> On Sat, 2010-01-09 at 10:14 -0500, Brian Walker wrote:
>    
>> This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):
>>
>> options ath9k debug=0xffffffff
>>
>> In config.mk I have:
>>
>> CONFIG_ATH9K_DEBUG=y
>> CONFIG_ATH9K_DEBUGFS=y
>>
>> Is there something else I need to do to get the debugging information
>> you're interested in?
>>
>> Brian
>>
>> On 01/07/2010 11:34 PM, Sujith wrote:
>>      
>>> Brian Walker wrote:
>>>
>>>        
>>>> Any update on getting this card working with the ath9k?
>>>>
>>>>          
>>> The logs that you posted earlier didn't contain debug information.
>>> Can you make sure that you load the driver with debug=0x601 and
>>> post the dmesg output on doing 'ifconfig wlan0 up' ?
>>>        
> I think the quoted message is pretty clear.  Maybe it would be better
> that you write under the quoted text so that you can see what you have
> quoted?
>
> debug=0xffffffff would likely produce too much information in the kernel
> log, so please change it to 0x601 as you were asked.
>    

Apologies for quoting after my messages. I will now quote before and 
then enter my response below the quotes.

Forgive me, but I am still confused. The only reason I set the 
debug=0xffffffff is that debug=0x601 did not seem to produce any 
debugging information in dmesg (or /var/log/messages). Even so, there 
does not appear to be any debugging information in either place with the 
debugging set so high. Do I need to blacklist the ath9k module so it 
does not load during boot and then manually modprobe it?

As I mentioned previously, if I rmmod ath9k, this shows in the kernel logs:

ath9k 0000:00:0a.0: PCI INT A disabled
ath9k: Driver unloaded

Then when I try modprobe ath9k debug=0x601, this shows in the kernel logs:

ath9k 0000:00:0a.0: failed to initialize device
ath9k 0000:00:0a.0: PCI INT A disabled
ath9k: probe of 0000:00:0a.0 failed with error -5

As one might imagine, it is quite frustrating to have to reboot just to 
load the ath9k module again.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):
> 
> options ath9k debug=0xffffffff
> 
> In config.mk I have:
> 
> CONFIG_ATH9K_DEBUG=y
> CONFIG_ATH9K_DEBUGFS=y

It should be:

CONFIG_ATH_DEBUG=y
CONFIG_ATH9K_DEBUGFS=y

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Sujith wrote:
> Brian Walker wrote:
>   
>> This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):
>>
>> options ath9k debug=0xffffffff
>>
>> In config.mk I have:
>>
>> CONFIG_ATH9K_DEBUG=y
>> CONFIG_ATH9K_DEBUGFS=y
>>     
>
> It should be:
>
> CONFIG_ATH_DEBUG=y
> CONFIG_ATH9K_DEBUGFS=y
>
> Sujith
>   

Ok, I will try rebuilding with that and rebooting. I believe I used 
CONFIG_ATH9K_DEBUG=y because of  the email from 12/17/2009:

>Can you build compat-wireless with debug enabled ?
>In config.mk, set CONFIG_ATH9K_DEBUG=y, rebuild and reload the driver
>with the module parameter "debug=0x600" and post the kernel log.
>
>Sujith


Thanks!

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/10/2010 12:12 AM, Sujith wrote:
> Brian Walker wrote:
>    
>> This is what I have in /etc/modprobe.d/local.conf (I run Fedora 11):
>>
>> options ath9k debug=0xffffffff
>>
>> In config.mk I have:
>>
>> CONFIG_ATH9K_DEBUG=y
>> CONFIG_ATH9K_DEBUGFS=y
>>      
> It should be:
>
> CONFIG_ATH_DEBUG=y
> CONFIG_ATH9K_DEBUGFS=y
>
> Sujith
>    

Ok! Now I have some debugging information. Here are the ath9k lines from 
dmesg right after boot:

ath9k 0000:00:0a.0: PCI INT A -> Link[LNKC] -> GSI 5 (level, low) -> IRQ 5
ath: UNDEFINED -> AWAKE
ath: serialize_regmode is 0
ath: Eeprom VER: 14, REV: 18
ath: EEPROM regdomain: 0x10
ath: EEPROM indicates we should expect a direct regpair map
ath: Country alpha2 being used: CO
ath: Regpair used: 0x10
ath: tx DMA: 512 buffers 1 desc/buf
ath: tx DMA map: f4ca0000 (76320) -> 0 (76320)
ath: beacon DMA: 4 buffers 1 desc/buf
ath: beacon DMA map: f4c5d000 (576) -> 0 (576)
ath: cachelsz 32 rxbufsize 3872
ath: rx DMA: 512 buffers 1 desc/buf
ath: rx DMA map: f4ce0000 (76320) -> 0 (76320)
phy0: Selected rate control algorithm 'ath9k_rate_control'
Registered led device: ath9k-phy0::radio
Registered led device: ath9k-phy0::assoc
Registered led device: ath9k-phy0::tx
phy0: Atheros AR9280 Rev:2 mem=0xfa5a0000, irq=5
cfg80211: Calling CRDA for country: CO
cfg80211: Regulatory domain changed to country: CO
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
     (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
cfg80211: Calling CRDA for country: US
cfg80211: Regulatory domain changed to country: US
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)

Here is dmesg after doing an "ifconfig wlan0 up":

ath: Starting driver with initial channel: 2412 MHz
ath: ah->misc_mode 0x4
ath: Attach a VIF of type: 2
ath: BSS Changed PREAMBLE 0
ath: BSS Changed CTS PROT 0
ath: Marking phy0 as idle
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset channel 4200038792 (2412Mhz) <7>ath: not-idle: 
enabling radio
ath: Set channel: 2412 MHz
ath: tx chmask: 1, rx chmask: 1
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: (2412 MHz) -> (2412 MHz), conf_is_ht40: 0
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset channel (2412 Mhz) reset status -22
ath: Unable to set channel
ath: Configure tx [queue/halq] [0/3],  aifs: 2, cw_min: 7, cw_max: 15, 
txop: 102
ath: Configure tx [queue/halq] [1/2],  aifs: 2, cw_min: 15, cw_max: 31, 
txop: 188
ath: Configure tx [queue/halq] [2/1],  aifs: 3, cw_min: 31, cw_max: 
1023, txop: 0
ath: Configure tx [queue/halq] [3/0],  aifs: 7, cw_min: 31, cw_max: 
1023, txop: 0
ath: Set HW RX filter: 0x2307
ath: Set HW RX filter: 0x2307
ADDRCONF(NETDEV_UP): wlan0: link is not ready
ath: Set HW RX filter: 0x2307

Here is dmesg after trying "iwlist wlan0 scan":

ath: Marking phy0 as not-idle
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset channel 4200038792 (2412Mhz) <7>ath: not-idle: 
enabling radio
ath: Set HW RX filter: 0x2117
ath: Set channel: 2412 MHz
ath: tx chmask: 3, rx chmask: 3
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: (2412 MHz) -> (2412 MHz), conf_is_ht40: 0
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed

These same type messages repeat for other channels.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Sujith]

Brian Walker wrote:
> Ok! Now I have some debugging information. Here are the ath9k lines from
> dmesg right after boot:
> 
> ath9k 0000:00:0a.0: PCI INT A -> Link[LNKC] -> GSI 5 (level, low) -> IRQ 5
> ath: UNDEFINED -> AWAKE
> ath: serialize_regmode is 0
> ath: Eeprom VER: 14, REV: 18
> ath: EEPROM regdomain: 0x10
> ath: EEPROM indicates we should expect a direct regpair map
> ath: Country alpha2 being used: CO
> ath: Regpair used: 0x10
> ath: tx DMA: 512 buffers 1 desc/buf
> ath: tx DMA map: f4ca0000 (76320) -> 0 (76320)
> ath: beacon DMA: 4 buffers 1 desc/buf
> ath: beacon DMA map: f4c5d000 (576) -> 0 (576)
> ath: cachelsz 32 rxbufsize 3872
> ath: rx DMA: 512 buffers 1 desc/buf
> ath: rx DMA map: f4ce0000 (76320) -> 0 (76320)
> phy0: Selected rate control algorithm 'ath9k_rate_control'
> Registered led device: ath9k-phy0::radio
> Registered led device: ath9k-phy0::assoc
> Registered led device: ath9k-phy0::tx
> phy0: Atheros AR9280 Rev:2 mem=0xfa5a0000, irq=5
> cfg80211: Calling CRDA for country: CO
> cfg80211: Regulatory domain changed to country: CO
>      (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>      (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
>      (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
>      (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
>      (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
> cfg80211: Calling CRDA for country: US
> cfg80211: Regulatory domain changed to country: US
>      (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>      (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
>      (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
>      (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>      (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>      (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>      (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)

Initialization is apparently okay - no errors here.

> Here is dmesg after doing an "ifconfig wlan0 up":
> 
> ath: Starting driver with initial channel: 2412 MHz
> ath: ah->misc_mode 0x4
> ath: Attach a VIF of type: 2
> ath: BSS Changed PREAMBLE 0
> ath: BSS Changed CTS PROT 0
> ath: Marking phy0 as idle
> ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 0x00000000
> ath: RTC stuck in MAC reset
> ath: Chip reset failed

The first HW reset fails as the contents of RTC_RC is weird.
Which probably means that the earlier initialization messed up somewhere,
corrupting the RTC register space - this would be my immediate guess.

Will take a look. Thanks for the debug log.

Sujith

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Wed, 2010-01-13 at 13:34 +0530, Sujith wrote:
> Brian Walker wrote:
> > Ok! Now I have some debugging information. Here are the ath9k lines from
> > dmesg right after boot:

That's for Ubiquiti SR71A:

# modprobe ath9k

ath9k 0000:04:01.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
ath: UNDEFINED -> AWAKE
ath: serialize_regmode is 1
ath: Eeprom VER: 14, REV: 22
ath: EEPROM regdomain: 0x0
ath: EEPROM indicates default country code should be used
ath: doing EEPROM country->regdmn map search
ath: country maps to regdmn code: 0x3a
ath: Country alpha2 being used: US
ath: Regpair used: 0x3a
ath: tx DMA: 512 buffers 1 desc/buf
ath: tx DMA map: ffff88003bfa0000 (80712) -> 3bfa0000 (80712)
ath: beacon DMA: 4 buffers 1 desc/buf
ath: beacon DMA map: ffff88003bf74000 (608) -> 3bf74000 (608)
ath: cachelsz 32 rxbufsize 3872
ath: rx DMA: 512 buffers 1 desc/buf
ath: rx DMA map: ffff88003bf40000 (80712) -> 3bf40000 (80712)
phy4: Selected rate control algorithm 'ath9k_rate_control'
Registered led device: ath9k-phy4::radio
Registered led device: ath9k-phy4::assoc
Registered led device: ath9k-phy4::tx
Registered led device: ath9k-phy4::rx
phy4: Atheros AR9280 Rev:2 mem=0xffffc90008320000, irq=19
udev: renamed network interface wlan0 to wlan4

# ifconfig wlan4 up

ath: Starting driver with initial channel: 2412 MHz
ath: ah->misc_mode 0x4
ath: timeout (100000 us) on reg 0x9860: 0xdeadbeef & 0x00000001 != 0x00000000
ath: Unable to reset hardware; reset status -5 (freq 2412 MHz)

> The first HW reset fails as the contents of RTC_RC is weird.
> Which probably means that the earlier initialization messed up somewhere,
> corrupting the RTC register space - this would be my immediate guess.
> 
> Will take a look. Thanks for the debug log.

I was able to make Ubiquiti SR71A work (at least it can scan now) by
treating it as an older Merlin 1.0 rather that Merlin 2.1 (version 0x80,
revision 2) as it claims to be:

diff --git a/drivers/net/wireless/ath/ath9k/hw.c b/drivers/net/wireless/ath/ath9k/hw.c
index 2311fe7..20f5375 100644
--- a/drivers/net/wireless/ath/ath9k/hw.c
+++ b/drivers/net/wireless/ath/ath9k/hw.c
@@ -690,7 +690,7 @@ static void ath9k_hw_init_mode_regs(struct ath_hw *ah)
 			ar9285PciePhy_clkreq_always_on_L1_9285,
 			ARRAY_SIZE(ar9285PciePhy_clkreq_always_on_L1_9285), 2);
 		}
-	} else if (AR_SREV_9280_20_OR_LATER(ah)) {
+	} else if (0) {
 		INIT_INI_ARRAY(&ah->iniModes, ar9280Modes_9280_2,
 			       ARRAY_SIZE(ar9280Modes_9280_2), 6);
 		INIT_INI_ARRAY(&ah->iniCommon, ar9280Common_9280_2,

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Wed, 2010-01-13 at 10:35 -0500, Pavel Roskin wrote:

> I was able to make Ubiquiti SR71A work (at least it can scan now) by
> treating it as an older Merlin 1.0 rather that Merlin 2.1 (version 0x80,
> revision 2) as it claims to be:

It has just occurred to me that both D-Link DWA-552 and Ubiquiti SR71A
are PCI devices (that is, not PCI Express).

I have a PCI Express device that is also version 0x80 revision 2, yet
it's working with ath9k (and with MadWifi since yesterday).

Maybe the code is assuming that Merlin 2.0 and newer is always PCI
Express?

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/13/2010 12:28 PM, Pavel Roskin wrote:
> On Wed, 2010-01-13 at 10:35 -0500, Pavel Roskin wrote:
>
>    
>> I was able to make Ubiquiti SR71A work (at least it can scan now) by
>> treating it as an older Merlin 1.0 rather that Merlin 2.1 (version 0x80,
>> revision 2) as it claims to be:
>>      
> It has just occurred to me that both D-Link DWA-552 and Ubiquiti SR71A
> are PCI devices (that is, not PCI Express).
>
> I have a PCI Express device that is also version 0x80 revision 2, yet
> it's working with ath9k (and with MadWifi since yesterday).
>
> Maybe the code is assuming that Merlin 2.0 and newer is always PCI
> Express?
>
>    

Thanks for continuing to work on this. My DWA-552 is indeed a PCI (not 
PCI-express) card. Perhaps the ath9k list of supported cards should be 
updated for the version of the DWA-552 that I have?

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

Quoting Brian Walker <walkerbm@charter.net>:

> Thanks for continuing to work on this. My DWA-552 is indeed a PCI (not
> PCI-express) card. Perhaps the ath9k list of supported cards should be
> updated for the version of the DWA-552 that I have?

The Wiki should be open for you to edit.  But I think it would be too  
premature to put new information into the Wiki before the scope of the  
problem is fully understood.

Some time ago I tested another Ubiquiti card, SR71-15, which is also  
miniPCI, and it worked fine with ath9k.  I'll need to retest it to see  
its revision.  I believe it's also AR9280.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/14/2010 01:15 AM, Pavel Roskin wrote:
> Quoting Brian Walker <walkerbm@charter.net>:
>
>> Thanks for continuing to work on this. My DWA-552 is indeed a PCI (not
>> PCI-express) card. Perhaps the ath9k list of supported cards should be
>> updated for the version of the DWA-552 that I have?
>
> The Wiki should be open for you to edit.  But I think it would be too 
> premature to put new information into the Wiki before the scope of the 
> problem is fully understood.
>
> Some time ago I tested another Ubiquiti card, SR71-15, which is also 
> miniPCI, and it worked fine with ath9k.  I'll need to retest it to see 
> its revision.  I believe it's also AR9280.
>

Ok. The only reason I purchased the DWA-552 was because it was listed at 
http://linuxwireless.org/en/users/Drivers/ath9k/products/external as 
supported by the ath9k driver. I am always very careful with hardware I 
buy to make sure it will work properly under Linux.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Thu, 2010-01-14 at 06:51 -0500, Brian Walker wrote:

> Ok. The only reason I purchased the DWA-552 was because it was listed at 
> http://linuxwireless.org/en/users/Drivers/ath9k/products/external as 
> supported by the ath9k driver. I am always very careful with hardware I 
> buy to make sure it will work properly under Linux.

However, it's just a bug in the driver.  Once the driver is fixed, your
device will be supported.  Wiki is not suitable for tracking bugs.  For
example, the MadWifi wiki is full of references to problems that no
longer exist in the code.  Nobody removed them from the wiki because it
would be just too much to ask from a developer who comes with a fix that
he also scans the wiki for all references to the problem being fixed.

If you document that your revision of DWA-552 is not supported, it will
be replicated elsewhere, and even if the next kernel will support your
card perfectly, somebody will still avoid buying DWA-552.

Using wiki to track bugs would discourage developers from touching it.
This happened to MadWifi, where the wiki is full of outdated and
unverifiable information, but I don't want it to happen to ath9k.

If you want to create a bug report, please use bugzilla.kernel.org.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Thu, 2010-01-14 at 18:13 -0500, Pavel Roskin wrote:

> If you want to create a bug report, please use bugzilla.kernel.org.

Maybe we could open a bug and then write something like this on the
ath9k page:

D-Link DWL-552 (revision X is affected by the bug NNN)
Ubiquiti SR71-12 (affected by the bug NNN)

This way, the users will be able to see if the bug has been fixed and in
which kernel revisions.  At the same time, the ath9k page won't be
littered with unmaintainable information.

The same rule could be used elsewhere in the Linux wireless wiki.  That
is, any references to bugs are only acceptable is accompanied with a
link to the bug tracker.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/13/2010 03:04 AM, Sujith wrote:
> Brian Walker wrote:
>    
>> Ok! Now I have some debugging information. Here are the ath9k lines from
>> dmesg right after boot:
>>
>> ath9k 0000:00:0a.0: PCI INT A ->  Link[LNKC] ->  GSI 5 (level, low) ->  IRQ 5
>> ath: UNDEFINED ->  AWAKE
>> ath: serialize_regmode is 0
>> ath: Eeprom VER: 14, REV: 18
>> ath: EEPROM regdomain: 0x10
>> ath: EEPROM indicates we should expect a direct regpair map
>> ath: Country alpha2 being used: CO
>> ath: Regpair used: 0x10
>> ath: tx DMA: 512 buffers 1 desc/buf
>> ath: tx DMA map: f4ca0000 (76320) ->  0 (76320)
>> ath: beacon DMA: 4 buffers 1 desc/buf
>> ath: beacon DMA map: f4c5d000 (576) ->  0 (576)
>> ath: cachelsz 32 rxbufsize 3872
>> ath: rx DMA: 512 buffers 1 desc/buf
>> ath: rx DMA map: f4ce0000 (76320) ->  0 (76320)
>> phy0: Selected rate control algorithm 'ath9k_rate_control'
>> Registered led device: ath9k-phy0::radio
>> Registered led device: ath9k-phy0::assoc
>> Registered led device: ath9k-phy0::tx
>> phy0: Atheros AR9280 Rev:2 mem=0xfa5a0000, irq=5
>> cfg80211: Calling CRDA for country: CO
>> cfg80211: Regulatory domain changed to country: CO
>>       (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>>       (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
>>       (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
>>       (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
>>       (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
>> cfg80211: Calling CRDA for country: US
>> cfg80211: Regulatory domain changed to country: US
>>       (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>>       (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
>>       (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
>>       (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>>       (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>>       (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>>       (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)
>>      
> Initialization is apparently okay - no errors here.
>
>    
>> Here is dmesg after doing an "ifconfig wlan0 up":
>>
>> ath: Starting driver with initial channel: 2412 MHz
>> ath: ah->misc_mode 0x4
>> ath: Attach a VIF of type: 2
>> ath: BSS Changed PREAMBLE 0
>> ath: BSS Changed CTS PROT 0
>> ath: Marking phy0 as idle
>> ath: timeout (100000 us) on reg 0x7000: 0xffffffff&  0x00000003 != 0x00000000
>> ath: RTC stuck in MAC reset
>> ath: Chip reset failed
>>      
> The first HW reset fails as the contents of RTC_RC is weird.
> Which probably means that the earlier initialization messed up somewhere,
> corrupting the RTC register space - this would be my immediate guess.
>
> Will take a look. Thanks for the debug log.
>
> Sujith
>    

Any update? I'd be glad to do further debugging with any code changes.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Sat, 2010-01-16 at 11:27 -0500, Brian Walker wrote:

> Any update? I'd be glad to do further debugging with any code changes.

It turns out that Ubiquiti SR71-15, which is also a miniPCI card with
AR9280 chipset is working properly.  It only supports the 5 GHz band.

Moreover, if I install both Ubiquiti SR71-15 and SR71-12, then SR71-12
works!  But if I modify ath9k to only match SR71-12, then it fails.

The debug messages are somewhat interesting:

mac80211: debugfs: failed to rename debugfs dir to netdev:wlan0
udev: renamed network interface wlan0 to wlan13
ath: timeout (100000 us) on reg 0x9860: 0xdeadbeef & 0x00000001 != 0x00000000
ath: Unable to reset hardware; reset status -5 (freq 2412 MHz)

0xdeadbeef actually comes from the register.  Either the Atheros chipset
is returning it (unlikely), or the driver is not reading it from the
actual hardware!

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Pavel Roskin wrote:
> On Sat, 2010-01-16 at 11:27 -0500, Brian Walker wrote:
>
>   
>> Any update? I'd be glad to do further debugging with any code changes.
>>     
>
> It turns out that Ubiquiti SR71-15, which is also a miniPCI card with
> AR9280 chipset is working properly.  It only supports the 5 GHz band.
>
> Moreover, if I install both Ubiquiti SR71-15 and SR71-12, then SR71-12
> works!  But if I modify ath9k to only match SR71-12, then it fails.
>
> The debug messages are somewhat interesting:
>
> mac80211: debugfs: failed to rename debugfs dir to netdev:wlan0
> udev: renamed network interface wlan0 to wlan13
> ath: timeout (100000 us) on reg 0x9860: 0xdeadbeef & 0x00000001 != 0x00000000
> ath: Unable to reset hardware; reset status -5 (freq 2412 MHz)
>
> 0xdeadbeef actually comes from the register.  Either the Atheros chipset
> is returning it (unlikely), or the driver is not reading it from the
> actual hardware!
>
>   
Any update? I see several new snapshots at 
http://wireless.kernel.org/download/compat-wireless-2.6/

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Mon, 2010-01-25 at 13:47 -0500, Brian Walker wrote:

> Any update?

Please test your card with this patch:
http://marc.info/?l=linux-wireless&m=126427347119805&w=2

My Ubiquiti SR71-12 is working with it.

>  I see several new snapshots at 
> http://wireless.kernel.org/download/compat-wireless-2.6/

That patch has not been committed yet.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/25/2010 02:34 PM, Pavel Roskin wrote:
> On Mon, 2010-01-25 at 13:47 -0500, Brian Walker wrote:
>
>    
>> Any update?
>>      
> Please test your card with this patch:
> http://marc.info/?l=linux-wireless&m=126427347119805&w=2
>
> My Ubiquiti SR71-12 is working with it.
>
>    
>>   I see several new snapshots at
>> http://wireless.kernel.org/download/compat-wireless-2.6/
>>      
> That patch has not been committed yet.
>
>    
Thanks for the pointer to the patch. I have applied it to the 2010-01-24 
version of compat-wireless, installed, and rebooted. Here is what is in 
dmesg after booting:

cfg80211: World regulatory domain updated:
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 5
PCI: setting IRQ 5 as level-triggered
ath9k 0000:00:0a.0: PCI INT A -> Link[LNKC] -> GSI 5 (level, low) -> IRQ 5
ath: UNDEFINED -> AWAKE
ath: serialize_regmode is 0
ath: Eeprom VER: 14, REV: 18
ath: EEPROM regdomain: 0x10
ath: EEPROM indicates we should expect a direct regpair map
ath: Country alpha2 being used: CO
ath: Regpair used: 0x10
ath: tx DMA: 512 buffers 1 desc/buf
ath: tx DMA map: f4c80000 (76320) -> 34c80000 (76320)
ath: beacon DMA: 4 buffers 1 desc/buf
ath: beacon DMA map: f4c3c000 (576) -> 34c3c000 (576)
ath: cachelsz 32 rxbufsize 3872
ath: rx DMA: 512 buffers 1 desc/buf
ath: rx DMA map: f4cc0000 (76320) -> 34cc0000 (76320)
phy0: Selected rate control algorithm 'ath9k_rate_control'
Registered led device: ath9k-phy0::radio
Registered led device: ath9k-phy0::assoc
Registered led device: ath9k-phy0::tx
Registered led device: ath9k-phy0::rx
phy0: Atheros AR9280 Rev:2 mem=0xfa3a0000, irq=5
cfg80211: Calling CRDA for country: CO
cfg80211: Regulatory domain changed to country: CO
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
     (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
cfg80211: Calling CRDA for country: US
cfg80211: Regulatory domain changed to country: US
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)

then after boot and doing "ifconfig wlan0 up"

ath: Starting driver with initial channel: 2412 MHz
ath: ah->misc_mode 0x4
ath: Attach a VIF of type: 2
ath: ah->misc_mode 0x4
ath: BSS Changed PREAMBLE 0
ath: BSS Changed CTS PROT 0
ath: Marking phy0 as idle
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset channel 4198050593 (2412Mhz) <7>ath: not-idle: 
enabling radio
ath: Set channel: 2412 MHz
ath: tx chmask: 1, rx chmask: 1
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Unable to stop TxDMA. Reset HAL!
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: (2412 MHz) -> (2412 MHz), conf_is_ht40: 0
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: RTC stuck in MAC reset
ath: Chip reset failed
ath: Unable to reset channel (2412 Mhz) reset status -22
ath: Unable to set channel
ath: Configure tx [queue/halq] [0/3],  aifs: 2, cw_min: 7, cw_max: 15, 
txop: 102
ath: Configure tx [queue/halq] [1/2],  aifs: 2, cw_min: 15, cw_max: 31, 
txop: 188
ath: Configure tx [queue/halq] [2/1],  aifs: 3, cw_min: 31, cw_max: 
1023, txop: 0
ath: Configure tx [queue/halq] [3/0],  aifs: 7, cw_min: 31, cw_max: 
1023, txop: 0
ath: Set HW RX filter: 0x2307
ath: Set HW RX filter: 0x2307
ADDRCONF(NETDEV_UP): wlan0: link is not ready
ath: Set HW RX filter: 0x2307

Those type messages repeat over and over after trying "iwlist wlan0 
scan". Also, about half the time I boot and try "iwconfig wlan0 up" the 
entire system just locks up. Other times the systems locks up (sometimes 
hours, sometimes days) later for seemingly no reason. Before adding the 
DWA-552 it had uptimes in excess of 100 days. I've taken to blacklisting 
the ath9k module unless I am testing a driver fix because this machine 
is my main firewall and fileserver.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Tue, 2010-01-26 at 19:52 -0500, Brian Walker wrote:

> ath: Marking phy0 as idle
> ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
> 0x00000000
> ath: RTC stuck in MAC reset
> ath: Chip reset failed

It looks like a completely different problem.  There is no 0xdeadbeef
involved and the failure is in a different place.

> ath: Unable to reset channel 4198050593 (2412Mhz) <7>ath: not-idle: 
> enabling radio

Apparently the statements to print "Unable to reset channel" are
incorrect in two places out of three in main.c.  ath_print() should be
annotated so that gcc would warn about invalid format.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 01/27/2010 06:13 PM, Pavel Roskin wrote:
> On Tue, 2010-01-26 at 19:52 -0500, Brian Walker wrote:
>
>    
>> ath: Marking phy0 as idle
>> ath: timeout (100000 us) on reg 0x7000: 0xffffffff&  0x00000003 !=
>> 0x00000000
>> ath: RTC stuck in MAC reset
>> ath: Chip reset failed
>>      
> It looks like a completely different problem.  There is no 0xdeadbeef
> involved and the failure is in a different place.
>
>    
>> ath: Unable to reset channel 4198050593 (2412Mhz)<7>ath: not-idle:
>> enabling radio
>>      
> Apparently the statements to print "Unable to reset channel" are
> incorrect in two places out of three in main.c.  ath_print() should be
> annotated so that gcc would warn about invalid format.
>
>    
Yes, the errors I am seeing appear to be the same ones I sent on January 
12. The ones on January 12 were from the 2009-12-11 codebase and the 
ones I sent yesterday are from the 2010-01-24 codebase. It would appear 
that whatever is wrong with my DWA-552 and the ath9k driver has not been 
addressed yet. Any idea when it might be? I would be glad to provide 
additional logs, etc. if it would help.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Pavel Roskin wrote:
> On Tue, 2010-01-26 at 19:52 -0500, Brian Walker wrote:
>
>   
>> ath: Marking phy0 as idle
>> ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
>> 0x00000000
>> ath: RTC stuck in MAC reset
>> ath: Chip reset failed
>>     
>
> It looks like a completely different problem.  There is no 0xdeadbeef
> involved and the failure is in a different place.
>
>   
>> ath: Unable to reset channel 4198050593 (2412Mhz) <7>ath: not-idle: 
>> enabling radio
>>     
>
> Apparently the statements to print "Unable to reset channel" are
> incorrect in two places out of three in main.c.  ath_print() should be
> annotated so that gcc would warn about invalid format.
>
>   
Is anyone looking at this? I have had this DWA-552 card since before 
Christmas and it is still not usable. I would be more than happy to test 
patches or provide more traces.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Fri, 2010-02-05 at 14:13 -0500, Brian Walker wrote:

> Is anyone looking at this? I have had this DWA-552 card since before 
> Christmas and it is still not usable. I would be more than happy to test 
> patches or provide more traces.

I ordered DWA-547, which is also reported to have problems, and I'm
going to look at the card when it arrives.  Unfortunately, the shipping
is taking too long.

If you give me the revision of your card, I could looks for it on eBay.
I would ask sellers if they have that revision.

I'm going to compare to registers accessed by ath9k and by the latest
Windows driver using ndiswrapper and the mmiotrace facility in the
kernel.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/05/2010 02:34 PM, Pavel Roskin wrote:
> On Fri, 2010-02-05 at 14:13 -0500, Brian Walker wrote:
>
>    
>> Is anyone looking at this? I have had this DWA-552 card since before
>> Christmas and it is still not usable. I would be more than happy to test
>> patches or provide more traces.
>>      
> I ordered DWA-547, which is also reported to have problems, and I'm
> going to look at the card when it arrives.  Unfortunately, the shipping
> is taking too long.
>
> If you give me the revision of your card, I could looks for it on eBay.
> I would ask sellers if they have that revision.
>
> I'm going to compare to registers accessed by ath9k and by the latest
> Windows driver using ndiswrapper and the mmiotrace facility in the
> kernel.
>
>    
Ok. What is the easiest way for me to determine the revision number of 
my card DWA-552? It was a recent purchase from NewEgg.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Fri, 2010-02-05 at 19:56 -0500, Brian Walker wrote:

> Ok. What is the easiest way for me to determine the revision number of 
> my card DWA-552? It was a recent purchase from NewEgg.

It's interesting that both reviews mentioning Linux give the card the
highest grade.

The revision should be written somewhere on the card or on the box.

I'm away from my DWA-552, but I looked for large images on the net, and
found one where the revision A1 is clearly seen at the end on P/N:
http://www.bcchardware.com/gallery/albums/D-LinkN/DWA_552Bundle.jpg

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/06/2010 12:20 AM, Pavel Roskin wrote:
> On Fri, 2010-02-05 at 19:56 -0500, Brian Walker wrote:
>
>    
>> Ok. What is the easiest way for me to determine the revision number of
>> my card DWA-552? It was a recent purchase from NewEgg.
>>      
> It's interesting that both reviews mentioning Linux give the card the
> highest grade.
>
> The revision should be written somewhere on the card or on the box.
>
> I'm away from my DWA-552, but I looked for large images on the net, and
> found one where the revision A1 is clearly seen at the end on P/N:
> http://www.bcchardware.com/gallery/albums/D-LinkN/DWA_552Bundle.jpg
>
>    

The sticker on my card says:

BWA552NA A2
HW Ver. A2 F/W Ver. 1.50

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Tue, 2010-02-09 at 20:00 -0500, Brian Walker wrote:

> The sticker on my card says:
> 
> BWA552NA A2
> HW Ver. A2 F/W Ver. 1.50

OK, I asked all sellers of reasonably priced DWA-552 on eBay about the
hardware revisions, but all answers so far have been "A1" or "I cannot
tell".

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Pavel Roskin wrote:
> On Tue, 2010-02-09 at 20:00 -0500, Brian Walker wrote:
>
>   
>> The sticker on my card says:
>>
>> BWA552NA A2
>> HW Ver. A2 F/W Ver. 1.50
>>     
>
> OK, I asked all sellers of reasonably priced DWA-552 on eBay about the
> hardware revisions, but all answers so far have been "A1" or "I cannot
> tell".
>
>   
Thanks for continuing to work on this. I got mine on NewEgg for ~$56 
plus shipping.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/10/2010 02:24 PM, Pavel Roskin wrote:
> On Tue, 2010-02-09 at 20:00 -0500, Brian Walker wrote:
>
>    
>> The sticker on my card says:
>>
>> BWA552NA A2
>> HW Ver. A2 F/W Ver. 1.50
>>      
> OK, I asked all sellers of reasonably priced DWA-552 on eBay about the
> hardware revisions, but all answers so far have been "A1" or "I cannot
> tell".
>
>    
Any update?

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Fri, 2010-02-19 at 22:00 -0500, Brian Walker wrote:

> Any update?

I got DWA-547 finally, but it worked with no problems.  It has AR5416,
so it's not surprising.  I also discovered that DWA-552 rev 1 that I
have is marked as DWA-547 on the PCB under the cover.  It's also AR5416.

There are two reviews for DWA-552 on newegg.com that mention Linux, and
both say that the card works.

I'm sorry, but I think I'll rather work with the hardware that I have
already.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/23/2010 04:30 PM, Pavel Roskin wrote:
> On Fri, 2010-02-19 at 22:00 -0500, Brian Walker wrote:
>
>    
>> Any update?
>>      
> I got DWA-547 finally, but it worked with no problems.  It has AR5416,
> so it's not surprising.  I also discovered that DWA-552 rev 1 that I
> have is marked as DWA-547 on the PCB under the cover.  It's also AR5416.
>
> There are two reviews for DWA-552 on newegg.com that mention Linux, and
> both say that the card works.
>
> I'm sorry, but I think I'll rather work with the hardware that I have
> already.
>
>    
Thanks for the update. That being the case, I think the documentation 
should be updated that the revision A2 DWA-552 with AR922X does NOT work 
with the ath9k driver currently. Perhaps one day I can use this card in 
Linux.

Brian

[ath9k-devel] hostapd-0.7-1 failed work with ath9k

[Zhang Yanfei-B21500]

Hi All

	I used the hostapd-0.7.1 and ath9k driver to setup a AP. With
ath9k, the station mode is work well. But in AP mode,  I follow the
guide the hostapd guide in ath9k. 

Enable the NL80211 driver and run hostapd, but it always failed as
below:
#hostapd -d /etc/hostapd.conf
Configuration file: /etc/hostapd.conf
Line 2: invalid/unknown driver 'nl80211'
1 errors found in configuration file '/etc/hostapd.conf'
root at wrt:/# cat /etc/hostapd.conf
interface=wlan0
driver=nl80211
ssid=test


Regards,
yanfei

> -----Original Message-----
> From: ath9k-devel-bounces at venema.h4ckr.net 
> [mailto:ath9k-devel-bounces at venema.h4ckr.net] On Behalf Of 
> Brian Walker
> Sent: Wednesday, February 24, 2010 9:22 AM
> To: Pavel Roskin
> Cc: ath9k-devel at venema.h4ckr.net
> Subject: Re: [ath9k-devel] trouble with DWA-552 and ath9k
> 
> On 02/23/2010 04:30 PM, Pavel Roskin wrote:
> > On Fri, 2010-02-19 at 22:00 -0500, Brian Walker wrote:
> >
> >    
> >> Any update?
> >>      
> > I got DWA-547 finally, but it worked with no problems.  It 
> has AR5416, 
> > so it's not surprising.  I also discovered that DWA-552 rev 
> 1 that I 
> > have is marked as DWA-547 on the PCB under the cover.  It's 
> also AR5416.
> >
> > There are two reviews for DWA-552 on newegg.com that mention Linux, 
> > and both say that the card works.
> >
> > I'm sorry, but I think I'll rather work with the hardware 
> that I have 
> > already.
> >
> >    
> Thanks for the update. That being the case, I think the 
> documentation should be updated that the revision A2 DWA-552 
> with AR922X does NOT work with the ath9k driver currently. 
> Perhaps one day I can use this card in Linux.
> 
> Brian
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> 
> 

[ath9k-devel] hostapd-0.7-1 failed work with ath9k

[Pavel Roskin]

On Wed, 2010-02-24 at 14:09 +0800, Zhang Yanfei-B21500 wrote:
> Hi All
> 
> 	I used the hostapd-0.7.1 and ath9k driver to setup a AP.
...
> > -----Original Message-----

Please don't hijack existing threads to discuss new topics.

-- 
Regards,
Pavel Roskin

[ath9k-devel] hostapd-0.7-1 failed work with ath9k

[yingqiang Ma]

Did you enable the nl80211 driver support when you compile the hostapd.

2010/2/24 Pavel Roskin <proski@gnu.org>

> On Wed, 2010-02-24 at 14:09 +0800, Zhang Yanfei-B21500 wrote:
> > Hi All
> >
> >       I used the hostapd-0.7.1 and ath9k driver to setup a AP.
> ...
> > > -----Original Message-----
>
> Please don't hijack existing threads to discuss new topics.
>
> --
> Regards,
> Pavel Roskin
> _______________________________________________
> 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/20100302/365f0f1e/attachment.htm 

[ath9k-devel] hostapd-0.7-1 failed work with ath9k

[Zhang Yanfei-B21500]

Sure i enabled the nl80211 driver for it.
 
Best Regards,
Yanfei Zhang
 


________________________________

	From: yingqiang Ma [mailto:yma.cool at gmail.com] 
	Sent: Tuesday, March 02, 2010 6:21 PM
	To: Pavel Roskin; Zhang Yanfei-B21500;
ath9k-devel at venema.h4ckr.net
	Subject: Re: [ath9k-devel] hostapd-0.7-1 failed work with ath9k
	
	
	Did you enable the nl80211 driver support when you compile the
hostapd.
	
	
	2010/2/24 Pavel Roskin <proski@gnu.org>
	

		On Wed, 2010-02-24 at 14:09 +0800, Zhang Yanfei-B21500
wrote:
		> Hi All
		>
		>       I used the hostapd-0.7.1 and ath9k driver to
setup a AP.
		
		...
		
		> > -----Original Message-----
		
		Please don't hijack existing threads to discuss new
topics.
		
		--
		Regards,
		Pavel Roskin
		_______________________________________________
		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/20100303/bc088b73/attachment.htm 

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Tue, 2010-02-23 at 20:21 -0500, Brian Walker wrote:

> Thanks for the update. That being the case, I think the documentation 
> should be updated that the revision A2 DWA-552 with AR922X does NOT work 
> with the ath9k driver currently. Perhaps one day I can use this card in 
> Linux.

I think the predominant result is positive, if the reviews on newegg.com
are considered and if they are about revision A2.

You still can try ndiswrapper with the Windows driver from atheros.cz.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/24/2010 01:40 AM, Pavel Roskin wrote:
> On Tue, 2010-02-23 at 20:21 -0500, Brian Walker wrote:
>
>    
>> Thanks for the update. That being the case, I think the documentation
>> should be updated that the revision A2 DWA-552 with AR922X does NOT work
>> with the ath9k driver currently. Perhaps one day I can use this card in
>> Linux.
>>      
> I think the predominant result is positive, if the reviews on newegg.com
> are considered and if they are about revision A2.
>
> You still can try ndiswrapper with the Windows driver from atheros.cz.
>
>    
We have no way of knowing which revision (A1 or A2) the newegg reviews 
are for. I have downloaded and tried the latest available 
compat-wireless ath9k driver for Linux and it still exibits the same 
problems I sent in earlier traces.

I know I could, but I have no interest in using ndiswrapper and a 
Windows driver.

Can anyone fix the problems that were exhibited in my previous traces 
for my revision A2 card, or should I just toss it in the trash and stay 
away from Atheros cards in the future?

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Wed, 2010-02-24 at 08:21 -0500, Brian Walker wrote:

> I know I could, but I have no interest in using ndiswrapper and a 
> Windows driver.

It could indicate if there is a hardware problem with your card or it's
a genuine problem with ath9k.

> Can anyone fix the problems that were exhibited in my previous traces 
> for my revision A2 card, or should I just toss it in the trash and stay 
> away from Atheros cards in the future?

I just checked all messages you posted in the list, and I don't see
kernel messages from ath9k anywhere.

You only posted the kernel messages before ath9k was even loaded:
https://lists.ath9k.org/pipermail/ath9k-devel/2009-December/002851.html

You were asked by Sujith to provide the kernel log when the interface is
brought up.  But your reply was empty:
https://lists.ath9k.org/pipermail/ath9k-devel/2009-December/002855.html

In other words, you have never provided the information you were asked
to provide.

I was wrong in assuming that you had the same problem as I had on
Ubiquiti SR71-12, so I didn't pay attention to the omission.  But now
that we know that your card is still unsupported, that information is
needed to deal with the problem.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Wed, 2010-02-24 at 13:09 -0500, Pavel Roskin wrote:

> I was wrong in assuming that you had the same problem as I had on
> Ubiquiti SR71-12, so I didn't pay attention to the omission.  But now
> that we know that your card is still unsupported, that information is
> needed to deal with the problem.

Sorry, I found it:
https://lists.ath9k.org/pipermail/ath9k-devel/2010-January/003019.html

Looking at it now.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Wed, 2010-02-24 at 13:14 -0500, Pavel Roskin wrote:

> Sorry, I found it:
> https://lists.ath9k.org/pipermail/ath9k-devel/2010-January/003019.html
> 
> Looking at it now.

I have bought DWA-552 from Newegg.  It's revision A2.  The chip is
marked as AR9223.

lspci should that it's exactly the same device as yours:

02:01.0 Network controller [0280]: Atheros Communications Inc. AR922X
Wireless Network Adapter [168c:0029] (rev 01)
        Subsystem: D-Link System Inc Device [1186:3a7a]
        Flags: bus master, 66MHz, medium devsel, latency 168, IRQ 18
        Memory at fe3f0000 (32-bit, non-prefetchable) [size=64K]
        Capabilities: [44] Power Management version 2
        Kernel driver in use: ath9k
        Kernel modules: ath9k

ath9h identifies the device as:
phy3: Atheros AR9280 Rev:2 mem=0xffffc90010260000, irq=18

I cannot reproduce your problem.  Everything is working just fine.

I checked what is read from the register 0x7000.  It's always 0x00000000
or 0x00000002 and never 0xffffffff.  I assume your card has some
hardware issue.

But the way, revision A2 doesn't just have another chip.  It looks very
different from the revision A1.  It's smaller and comes with an
additional low-profile bracket.  Antennas are closer to each other.  The
picture on the box shows it clearly.  D-Link must be getting really low
on model numbers if they are reusing the same number for a substantially
different product.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Peter Stuge]

Pavel Roskin wrote:
> D-Link must be getting really low on model numbers if they are
> reusing the same number for a substantially different product.

I think this is pretty common for a few years already.

Horrible, but common. :\


//Peter

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/25/2010 01:31 PM, Pavel Roskin wrote:
> On Wed, 2010-02-24 at 13:14 -0500, Pavel Roskin wrote:
>
>    
>> Sorry, I found it:
>> https://lists.ath9k.org/pipermail/ath9k-devel/2010-January/003019.html
>>
>> Looking at it now.
>>      
> I have bought DWA-552 from Newegg.  It's revision A2.  The chip is
> marked as AR9223.
>
> lspci should that it's exactly the same device as yours:
>
> 02:01.0 Network controller [0280]: Atheros Communications Inc. AR922X
> Wireless Network Adapter [168c:0029] (rev 01)
>          Subsystem: D-Link System Inc Device [1186:3a7a]
>          Flags: bus master, 66MHz, medium devsel, latency 168, IRQ 18
>          Memory at fe3f0000 (32-bit, non-prefetchable) [size=64K]
>          Capabilities: [44] Power Management version 2
>          Kernel driver in use: ath9k
>          Kernel modules: ath9k
>
> ath9h identifies the device as:
> phy3: Atheros AR9280 Rev:2 mem=0xffffc90010260000, irq=18
>
> I cannot reproduce your problem.  Everything is working just fine.
>
> I checked what is read from the register 0x7000.  It's always 0x00000000
> or 0x00000002 and never 0xffffffff.  I assume your card has some
> hardware issue.
>
> But the way, revision A2 doesn't just have another chip.  It looks very
> different from the revision A1.  It's smaller and comes with an
> additional low-profile bracket.  Antennas are closer to each other.  The
> picture on the box shows it clearly.  D-Link must be getting really low
> on model numbers if they are reusing the same number for a substantially
> different product.
>
>    
You do indeed appear to have the exact same DWA-552 as me. Is it 
possible that the problem is with the way my older PIII AX63Pro 
motherboard is interacting with the card? The only difference I see is 
the latency on mine is 32 and not 168. Unfortunately, I believe it has 
been too long (and I no longer have the box) to return this card.

What version of compat-wireless are you using and with what kernel and 
on what distro?

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Thu, 2010-02-25 at 20:10 -0500, Brian Walker wrote:

> You do indeed appear to have the exact same DWA-552 as me. Is it 
> possible that the problem is with the way my older PIII AX63Pro 
> motherboard is interacting with the card? The only difference I see is 
> the latency on mine is 32 and not 168. Unfortunately, I believe it has 
> been too long (and I no longer have the box) to return this card.

It can easily be the latency.  You can see the BIOS settings.  Some
motherboards allow changing the PCI latency.  Try increasing it.

I tried decreasing the PCI latency using setpci, but it didn't appear to
work.

> What version of compat-wireless are you using and with what kernel and 
> on what distro?

I'm using Fedora 11 with a kernel from the wireless-testing repository
(currently identifying itself as "2.6.33-wl").

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/26/2010 06:35 PM, Pavel Roskin wrote:
> On Thu, 2010-02-25 at 20:10 -0500, Brian Walker wrote:
>
>    
>> You do indeed appear to have the exact same DWA-552 as me. Is it
>> possible that the problem is with the way my older PIII AX63Pro
>> motherboard is interacting with the card? The only difference I see is
>> the latency on mine is 32 and not 168. Unfortunately, I believe it has
>> been too long (and I no longer have the box) to return this card.
>>      
> It can easily be the latency.  You can see the BIOS settings.  Some
> motherboards allow changing the PCI latency.  Try increasing it.
>
> I tried decreasing the PCI latency using setpci, but it didn't appear to
> work.
>
>    
>> What version of compat-wireless are you using and with what kernel and
>> on what distro?
>>      
> I'm using Fedora 11 with a kernel from the wireless-testing repository
> (currently identifying itself as "2.6.33-wl").
>
>    
Any chance you could try testing with a stock kernel and one of the 
compat-wireless snapshots from 
http://wireless.kernel.org/download/compat-wireless-2.6/? Also, are we 
to just assume that at this point I have a hardware problem and it 
couldn't be ath9k?

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Mon, 2010-03-01 at 19:48 -0500, Brian Walker wrote:

> Any chance you could try testing with a stock kernel and one of the 
> compat-wireless snapshots from 
> http://wireless.kernel.org/download/compat-wireless-2.6/?

I tested the card with the current Fedora 11 kernel,
2.6.30.10-105.2.23.fc11.x86_64

I also tried that kernel with compat-wireless-2.6.33 and
compat-wireless-2010-03-02.  The card was working fine in all cases.

> Also, are we 
> to just assume that at this point I have a hardware problem and it 
> couldn't be ath9k?

Yes.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 02/26/2010 06:35 PM, Pavel Roskin wrote:
> On Thu, 2010-02-25 at 20:10 -0500, Brian Walker wrote:
>
>    
>> You do indeed appear to have the exact same DWA-552 as me. Is it
>> possible that the problem is with the way my older PIII AX63Pro
>> motherboard is interacting with the card? The only difference I see is
>> the latency on mine is 32 and not 168. Unfortunately, I believe it has
>> been too long (and I no longer have the box) to return this card.
>>      
> It can easily be the latency.  You can see the BIOS settings.  Some
> motherboards allow changing the PCI latency.  Try increasing it.
>
> I tried decreasing the PCI latency using setpci, but it didn't appear to
> work.
>
>    
>> What version of compat-wireless are you using and with what kernel and
>> on what distro?
>>      
> I'm using Fedora 11 with a kernel from the wireless-testing repository
> (currently identifying itself as "2.6.33-wl").
>
>    
Just used setpci -v -d *:* latency_timer=b0 to set the latency of all 
devices to 176, did a modprobe ath9k and an ifconfig wlan0 up. System 
didn't crash, but still getting:

ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Unable to stop TxDMA. Reset HAL!
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset channel (2412 Mhz) reset status -22
ath: Unable to set channel

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Pavel Roskin]

On Mon, 2010-03-01 at 19:55 -0500, Brian Walker wrote:

> Just used setpci -v -d *:* latency_timer=b0 to set the latency of all 
> devices to 176, did a modprobe ath9k and an ifconfig wlan0 up. System 
> didn't crash, but still getting:
> 
> ath: Failed to stop TX DMA in 100 msec after killing last frame
> ath: Unable to stop TxDMA. Reset HAL!
> ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
> 0x00000000

I'm afraid the only sure way to set PCI latency is from BIOS.

-- 
Regards,
Pavel Roskin

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

On 03/02/2010 11:33 PM, Pavel Roskin wrote:
> On Mon, 2010-03-01 at 19:55 -0500, Brian Walker wrote:
>
>    
>> Just used setpci -v -d *:* latency_timer=b0 to set the latency of all
>> devices to 176, did a modprobe ath9k and an ifconfig wlan0 up. System
>> didn't crash, but still getting:
>>
>> ath: Failed to stop TX DMA in 100 msec after killing last frame
>> ath: Unable to stop TxDMA. Reset HAL!
>> ath: timeout (100000 us) on reg 0x7000: 0xffffffff&  0x00000003 !=
>> 0x00000000
>>      
> I'm afraid the only sure way to set PCI latency is from BIOS.
>
>    
My motherboard is a quite old (~11 years) Acer AX63 Pro. It does not 
have any options in the bios to set the PCI latency. I am still 
experiencing the following errors in my kernel logs:

Generic kernel compatibility enabled based on linux-next next-20100113
cfg80211: Calling CRDA to update world regulatory domain
cfg80211: World regulatory domain updated:
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
PCI: setting IRQ 5 as level-triggered
ath9k 0000:00:0a.0: found PCI INT A -> IRQ 5
ath: EEPROM regdomain: 0x10
ath: EEPROM indicates we should expect a direct regpair map
ath: Country alpha2 being used: CO
ath: Regpair used: 0x10
phy0: Selected rate control algorithm 'ath9k_rate_control'
cfg80211: Calling CRDA for country: CO
Registered led device: ath9k-phy0::radio
Registered led device: ath9k-phy0::assoc
Registered led device: ath9k-phy0::tx
egistered led device: ath9k-phy0::rx
phy0: Atheros AR9280 Rev:2 mem=0xfae00000, irq=5
cfg80211: Regulatory domain changed to country: CO
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 20000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 20000 KHz), (300 mBi, 2300 mBm)
     (5735000 KHz - 5835000 KHz @ 20000 KHz), (300 mBi, 3000 mBm)
cfg80211: Calling CRDA for country: US
cfg80211: Regulatory domain changed to country: US
     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
     (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset channel (2412 MHz), reset status -22
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Unable to stop TxDMA. Reset HAL!
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset channel (2412 MHz), reset status -22
ath: Unable to set channel
ADDRCONF(NETDEV_UP): wlan0: link is not ready
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset channel (2412 MHz), reset status -22
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Unable to stop TxDMA. Reset HAL!
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22
ath: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset channel (2412 MHz), reset status -22
ath: Unable to set channel
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Failed to stop TX DMA in 100 msec after killing last frame
ath: Unable to stop TxDMA. Reset HAL!
ath: timeout (100000 us) on reg 0x7000: 0xffffffff & 0x00000003 != 
0x00000000
ath: Chip reset failed
ath: Unable to reset hardware; reset status -22

Since this appears to work for others with kernel 
2.6.30.10-105.fc11.i686.PAE and wireless snapshot 
compat-wireless-2010-03-07 I can only come to the conclusion that either 
this card is incompatible with my motherboard or has a hardware problem. 
Caveat emptor I suppose? As such, I'd recommend staying away from the A2 
version of the DWA-552. Actually, I suggest staying away from all cards 
from companies that reuse the same model number for cards with 
completely different wireless chipsets.

Brian

[ath9k-devel] trouble with DWA-552 and ath9k

[Brian Walker]

Probably should have also sent the lspci output for the DWA-552:

00:0a.0 Network controller [0280]: Atheros Communications Inc. AR922X 
Wireless Network Adapter [168c:0029] (rev 01)
        Subsystem: D-Link System Inc Device [1186:3a7a]
        Flags: 66MHz, medium devsel, IRQ 5
        [virtual] Memory at f8040000 (32-bit, non-prefetchable) [size=64K]
        Capabilities: [44] Power Management version 2
        Kernel driver in use: ath9k
        Kernel modules: ath9k


Brian