[RFT] ath9k_htc - please test new images

[RFT] ath9k_htc - please test new images

[Adrian Chadd]

Hi all,

I have a couple of new test images for ath9k_htc users to test.

http://developer.atheros.com/~adrian/htc/20130222/

These images are built by the open source tensilica toolchain rather
than the proprietary toolchain.

Their version id has been bumped from 1.3 to 1.4.

I'd appreciate any testing people can do.

Thanks!



Adrian

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Adrian Chadd]

Hi all,

I have a couple of new test images for ath9k_htc users to test.

http://developer.atheros.com/~adrian/htc/20130222/

These images are built by the open source tensilica toolchain rather
than the proprietary toolchain.

Their version id has been bumped from 1.3 to 1.4.

I'd appreciate any testing people can do.

Thanks!



Adrian

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Oleksij Rempel]

Am 22.02.2013 20:33, schrieb Adrian Chadd:
> Hi all,
>
> I have a couple of new test images for ath9k_htc users to test.
>
> http://developer.atheros.com/~adrian/htc/20130222/
>
> These images are built by the open source tensilica toolchain rather
> than the proprietary toolchain.
>
> Their version id has been bumped from 1.3 to 1.4.
>
> I'd appreciate any testing people can do.
>
> Thanks!

Hi Adrian,

current driver has this check:
htc_drv_init.c:	    priv->fw_version_minor != MINOR_VERSION_REQ) {
htc_drv_init.c:			MAJOR_VERSION_REQ, MINOR_VERSION_REQ);

so it wont accept 1.4. Currently i compiling kernel without this 
restriction. Then i'll be able to test it.

-- 
Regards,
Oleksij

Re: [ath9k-devel] [RFT] ath9k_htc - please test new images

[Adrian Chadd]

On 23 February 2013 01:54, Oleksij Rempel <bug-track@fisher-privat.net> wrote:
>
> Hi Adrian,
>
> current driver has this check:
> htc_drv_init.c:     priv->fw_version_minor != MINOR_VERSION_REQ) {
> htc_drv_init.c:                 MAJOR_VERSION_REQ, MINOR_VERSION_REQ);
>
> so it wont accept 1.4. Currently i compiling kernel without this
> restriction. Then i'll be able to test it.

Oops! Guess I shouldn't have bumped the version numbers :-)

Thanks,



Adrian

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Adrian Chadd]

On 23 February 2013 01:54, Oleksij Rempel <bug-track@fisher-privat.net> wrote:
>
> Hi Adrian,
>
> current driver has this check:
> htc_drv_init.c:     priv->fw_version_minor != MINOR_VERSION_REQ) {
> htc_drv_init.c:                 MAJOR_VERSION_REQ, MINOR_VERSION_REQ);
>
> so it wont accept 1.4. Currently i compiling kernel without this
> restriction. Then i'll be able to test it.

Oops! Guess I shouldn't have bumped the version numbers :-)

Thanks,



Adrian

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Oleksij Rempel]

Am 23.02.2013 12:41, schrieb Adrian Chadd:
> On 23 February 2013 01:54, Oleksij Rempel <bug-track@fisher-privat.net> wrote:
>>
>> Hi Adrian,
>>
>> current driver has this check:
>> htc_drv_init.c:     priv->fw_version_minor != MINOR_VERSION_REQ) {
>> htc_drv_init.c:                 MAJOR_VERSION_REQ, MINOR_VERSION_REQ);
>>
>> so it wont accept 1.4. Currently i compiling kernel without this
>> restriction. Then i'll be able to test it.
>
> Oops! Guess I shouldn't have bumped the version numbers :-)

Now results:
- Current wireless-testing.git is broken for my htc_* devices. none of 
firmwares is working.
- I use 3.8.0-rc7 from wireless-testing.git for testing.
- htc_7010.fw is working.
- htc_9271.fw do not working. it fails on authentication:

[32305.187783] usb 3-2: Product: UB91C
[32305.187789] usb 3-2: Manufacturer: ATHEROS
[32305.187794] usb 3-2: SerialNumber: 12345
[32305.262708] usb 3-2: ath9k_htc: Firmware htc_9271.fw requested
[32305.263302] usbcore: registered new interface driver ath9k_htc
[32305.526943] usb 3-2: ath9k_htc: Transferred FW: htc_9271.fw, size: 48820
[32305.763378] ath9k_htc 3-2:1.0: ath9k_htc: HTC initialized with 33 credits
[32306.076930] ath9k_htc 3-2:1.0: ath9k_htc: FW Version: 1.4
[32306.076939] ath: EEPROM regdomain: 0x0
[32306.076943] ath: EEPROM indicates default country code should be used
[32306.076946] ath: doing EEPROM country->regdmn map search
[32306.076950] ath: country maps to regdmn code: 0x3a
[32306.076954] ath: Country alpha2 being used: US
[32306.076957] ath: Regpair used: 0x3a
[32306.088331] ieee80211 phy6: Atheros AR9271 Rev:1
[32306.573836] IPv6: ADDRCONF(NETDEV_UP): wlan2: link is not ready
[32309.755882] wlan2: authenticate with 74:31:70:62:d3:cf
[32310.183668] wlan2: send auth to 74:31:70:62:d3:cf (try 1/3)
[32310.185673] wlan2: authenticated
[32310.187193] wlan2: associate with 74:31:70:62:d3:cf (try 1/3)
[32310.190965] wlan2: RX AssocResp from 74:31:70:62:d3:cf (capab=0x411 
status=0 aid=4)
[32310.224419] wlan2: associated
[32310.224448] IPv6: ADDRCONF(NETDEV_CHANGE): wlan2: link becomes ready
[32320.218444] wlan2: disassociating from 74:31:70:62:d3:cf by local 
choice (reason=3)
[32320.692076] wlan2: deauthenticating from 74:31:70:62:d3:cf by local 
choice (reason=3)

-- 
Regards,
Oleksij

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Felix Fietkau]

On 2013-02-23 9:13 PM, Oleksij Rempel wrote:
> Am 23.02.2013 12:41, schrieb Adrian Chadd:
>> On 23 February 2013 01:54, Oleksij Rempel <bug-track@fisher-privat.net> wrote:
>>>
>>> Hi Adrian,
>>>
>>> current driver has this check:
>>> htc_drv_init.c:     priv->fw_version_minor != MINOR_VERSION_REQ) {
>>> htc_drv_init.c:                 MAJOR_VERSION_REQ, MINOR_VERSION_REQ);
>>>
>>> so it wont accept 1.4. Currently i compiling kernel without this
>>> restriction. Then i'll be able to test it.
>>
>> Oops! Guess I shouldn't have bumped the version numbers :-)
> 
> Now results:
> - Current wireless-testing.git is broken for my htc_* devices. none of 
> firmwares is working.
> - I use 3.8.0-rc7 from wireless-testing.git for testing.
> - htc_7010.fw is working.
> - htc_9271.fw do not working. it fails on authentication:
Same here (AR9271).

- Felix

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Adrian Chadd]

On 23 February 2013 12:47, Felix Fietkau <nbd@openwrt.org> wrote:

>> - I use 3.8.0-rc7 from wireless-testing.git for testing.
>> - htc_7010.fw is working.
>> - htc_9271.fw do not working. it fails on authentication:
> Same here (AR9271).

Right. http://developer.atheros.com/~adrian/htc/ has another snapshot
from november 2012 - would you mind checking if that firmware works on
the AR9271, same driver?
(but with the revision changed back, obviously.)

Thanks,


Adrian

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Oleksij Rempel]

Am 23.02.2013 22:05, schrieb Adrian Chadd:
> On 23 February 2013 12:47, Felix Fietkau <nbd@openwrt.org> wrote:
>
>>> - I use 3.8.0-rc7 from wireless-testing.git for testing.
>>> - htc_7010.fw is working.
>>> - htc_9271.fw do not working. it fails on authentication:
>> Same here (AR9271).
>
> Right. http://developer.atheros.com/~adrian/htc/ has another snapshot
> from november 2012 - would you mind checking if that firmware works on
> the AR9271, same driver?
> (but with the revision changed back, obviously.)

snapshot from november 2012 is working.


-- 
Regards,
Oleksij

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Felix Fietkau]

On 2013-02-23 10:05 PM, Adrian Chadd wrote:
> On 23 February 2013 12:47, Felix Fietkau <nbd@openwrt.org> wrote:
> 
>>> - I use 3.8.0-rc7 from wireless-testing.git for testing.
>>> - htc_7010.fw is working.
>>> - htc_9271.fw do not working. it fails on authentication:
>> Same here (AR9271).
> 
> Right. http://developer.atheros.com/~adrian/htc/ has another snapshot
> from november 2012 - would you mind checking if that firmware works on
> the AR9271, same driver?
> (but with the revision changed back, obviously.)
That one works.

- Felix

[ath9k-devel] [RFT] ath9k_htc - please test new images

[Adrian Chadd]

On 23 February 2013 13:13, Felix Fietkau <nbd@openwrt.org> wrote:

>> Right. http://developer.atheros.com/~adrian/htc/ has another snapshot
>> from november 2012 - would you mind checking if that firmware works on
>> the AR9271, same driver?
>> (but with the revision changed back, obviously.)
>
> That one works.

cool, thanks.

The only that's changed in the driver codebase is the compiler being
used and the specific tensilica CPU profile (and thus automatically
generated header files.)

I'll poke the development list to see if the people hacking on this
can figure out what isn't working right with the AR9271.

Thanks for testing!



Adrian