Re: ath9k: AR9462 btcoex doesn't appear to work

[Sherwin Soltani <sherwin@wybc.com>]

On 08/25/2012 09:58 PM, Sujith Manoharan wrote:
> Sherwin Soltani wrote:
>> I have an Acer Aspire V5-571 which contains the AR9462 chipset (PCI ID
>> 168c:0034), supported by the ath9k driver. This chipset claims to
>> support Bluetooth coexistence. However, after booting up, my kernel does
>> not register any Bluetooth devices.
>>
>> I am using kernel version 3.6.0-999-generic (apw@gomeisa) (gcc version
>> 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)). I also tried 3.4 and 3.2 kernels
>> without success. Additionally, I can try to do
>>
>> rmmod ath9k && modprobe ath9k btcoex_enable=1
>>
>> In the message log, I can see this:
>>
>> [ 1204.643120] ath9k: ath9k: Driver unloaded
>> [ 1214.203099] ath: phy1: Disabling ASPM since BTCOEX is enabled
>> [ 1214.203105] ath: EEPROM regdomain: 0x6a
>> [ 1214.203106] ath: EEPROM indicates we should expect a direct regpair map
>> [ 1214.203108] ath: Country alpha2 being used: 00
>> [ 1214.203109] ath: Regpair used: 0x6a
>>
>> However 'hcitool dev' shows no devices after this operation.
> What does lsusb -v show ?

It shows:

03:00.0 Network controller: Atheros Communications Inc. AR9462 Wireless 
Network Adapter (rev 01)
     Subsystem: Lite-On Communications Inc Device 6621
     Flags: bus master, fast devsel, latency 0, IRQ 19
     Memory at f0500000 (64-bit, non-prefetchable) [size=512K]
     Expansion ROM at dfa00000 [disabled] [size=64K]
     Capabilities: [40] Power Management version 2
     Capabilities: [50] MSI: Enable- Count=1/4 Maskable+ 64bit+
     Capabilities: [70] Express Endpoint, MSI 00
     Capabilities: [100] Advanced Error Reporting
     Capabilities: [140] Virtual Channel
     Capabilities: [160] Device Serial Number 00-00-00-00-00-00-00-00
     Kernel driver in use: ath9k
     Kernel modules: ath9k

> ...
>
> Reload ath3k and something like this should be seen:
>
> root@linux-test ~# hciconfig hci1 up
> root@linux-test ~# hcitool dev
> Devices:
>          hci1    00:03:7F:44:84:8A

By ath3k, I assume you meant ath9k since that is the driver I was using. 
After loading up the new firmware, I still have no devices shown (and 
'hciconfig hci1 up' replies with 'No such device'). Rebooting doesn't help.

Here is what I see in the message buffer after re-inserting ath9k:
[19157.961388] ath9k: ath9k: Driver unloaded
[19178.196366] ath: phy2: ASPM enabled: 0x43
[19178.196371] ath: EEPROM regdomain: 0x6a
[19178.196372] ath: EEPROM indicates we should expect a direct regpair map
[19178.196375] ath: Country alpha2 being used: 00
[19178.196376] ath: Regpair used: 0x6a

... lots of cfg80211 info ...

[19178.197491] Registered led device: ath9k-phy2
[19178.197500] ieee80211 phy2: Atheros AR9462 Rev:2 
mem=0xffffc90016300000, irq=19
[19178.211376] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready
[19186.703984] wlan0: authenticate with 2c:b0:5d:45:33:58
[19186.708760] wlan0: send auth to 2c:b0:5d:45:33:58 (try 1/3)
[19186.759145] wlan0: authenticated
[19186.765888] wlan0: associate with 2c:b0:5d:45:33:58 (try 1/3)
[19186.863455] wlan0: RX AssocResp from 2c:b0:5d:45:33:58 (capab=0x411 
status=0 aid=2)
[19186.863561] wlan0: associated

> Coexistence testing is ongoing and bugs are still being ironed out,
> but basic support is mostly done.
>
>> Aside from the Bluetooth problems, there are also problems with
>> frequently dropping signals (even when I am very close to the access
>> point), but right now that's a secondary concern to Bluetooth for me.
> Can you try with the latest driver (using compat-wireless) and see if the
> problem persists ? If it does, please post the kernel log with driver
> debugging set to 0x8f49.

I will take a look at what happens with compat-wireless.