compat-wireless-2014-05-22 breaks ath10k in an Archer C7 v2

Michal Kazior michal.kazior at tieto.com
Thu Jun 5 05:20:11 PDT 2014


On 5 June 2014 13:49, Sven Schnelle <svens at stackframe.org> wrote:
> On 06/05/2014 01:36 PM, Kalle Valo wrote:
>> Sven Schnelle <svens at stackframe.org> writes:
[...]
>>> I see the following output on the firmware console:
>>>
>>> otp main...
>>> Table initialized..
>>> streamCnt 0
>>> Golden Bin/fake Board Template used ...
>>> otp done...
>>> assertion failed? pc=0x948f19, line=5555, dump area=0x40ad04
>>> Target ID: 0x4100016c (1090519404)
>>> Debug Info:
>>> 0x4100016c 0x000015b3 0x00948f19 0x00955b31
>>> 0x00948f19 0x00060530 0x00000004 0x00000000
>>> 0x00000000 0x004045f0 0x00404610 0x00955a00
>>> 0x0000
>>>
>>> However, the AP where OTP now fails is an engineering sample,
>>> and probably doesn't contain any valid calibration. It did work
>>> in the past, but with the latest ath10k the driver doesn't load
>>> because of OTP failing. So maybe that output is helpful here.
>>
>> What we could do is add 'skip_otp' (or similar) module parameter,
>> otp.bin is not run and then you just get the default calibration data.
>>
> Looking at the backtrace, can you tell whether the calibration data is
> really
> the issue? I tried now with the firmware-2.bin and board.bin files from
> https://github.com/kvalo/ath10k-firmware/ath10k/QCA988X/hw2.0/ and
> i'm getting the same error.

This seems to confirm otp fails to load with error code 2.

The assert stack trace points to copy engine in ROM image. Wierd.

Anyway, I've extracted firmware files from TP-Link Archer C7 original
rootfs. Can you give it a try, please? See the OpenWRT ticket:

  https://dev.openwrt.org/ticket/16589


Michał



More information about the ath10k mailing list