QCA9880/9890 rfkill

[William Riba]

That unfortunately seems to be the consensus. At least one of our card
vendors believes it's a firmware and not an EEPROM setting issue (waiting to
hear from the other). They're attempting to resolve it through Qualcomm.  I
have doubts there's going to be a resolution any time soon... Thanks much
for the advice!

-bill


-----Original Message-----
From: Ben Greear [mailto:greearb at candelatech.com] 
Sent: Wednesday, September 23, 2015 3:37 PM
To: Adrian Chadd; William Riba
Cc: ath10k at lists.infradead.org
Subject: Re: QCA9880/9890 rfkill

At least 10.1.467 era firmware ignores some of the OTP settings, so you may
have to go poking in the firmware source to be certain of what can be
configured and/or fix the limitations.

Thanks,
Ben

On 09/23/2015 01:25 PM, Adrian Chadd wrote:
> iirc otp.bin is the image used when you don't have one; it should be 
> coming out of OTP or on-board flash.
>
>
> -a
>
>
> On 23 September 2015 at 12:32, William Riba <wriba at wriba.com> wrote:
>> Hi Adrian - thank you for the quick response. I assume you're referring
the otp.bin firmware file? Is there a mapping of the OTP to EEPROM settings
available? That would be an easy fix if I can make a modification there to
activate RFKILL.
>>
>> -bill
>>
>>
>>
>>
>> -----Original Message-----
>> From: adrian.chadd at gmail.com [mailto:adrian.chadd at gmail.com] On 
>> Behalf Of Adrian Chadd
>> Sent: Wednesday, September 23, 2015 1:58 PM
>> To: William Riba
>> Cc: ath10k at lists.infradead.org
>> Subject: Re: QCA9880/9890 rfkill
>>
>> Hi,
>>
>> The GPIO line needs to be hooked in via the relevant EEPROM/OTP setting
so the firmware programs things in correctly.
>>
>> Go dump the OTP and see what the rfkill firmware line is.
>>
>>
>>
>> -adrian
>>
>>
>> On 23 September 2015 at 11:55, William Riba <wriba at wriba.com> wrote:
>>> Does anybody know whether rfkill in its hard block flavor is 
>>> supported in any QCA9880/9890 based device and/or firmware load? 
>>> While vendors refer to the WiFi disable/GPIO line in their reference 
>>> schematics, it is not functional.  What I see after testing multiple 
>>> cards is if the line is asserted during operation, there seems to be 
>>> no effect. If it is asserted at boot-up, the device does not enumerate
on the PCI bus.
>>>
>>> If Qualcomm does not intend to support this functionality anymore it 
>>> would be nice to know. Hard kills are a regulatory requirement for 
>>> certain applications. While we would like to use a QCA9890 in a new 
>>> application, there's no way it could make it through qualifications 
>>> with its current operation.  Any additional information on this would be
welcome. Thanks.
>>>
>>>
>>> _______________________________________________
>>> ath10k mailing list
>>> ath10k at lists.infradead.org
>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>
>
> _______________________________________________
> ath10k mailing list
> ath10k at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k
>


--
Ben Greear <greearb at candelatech.com>
Candela Technologies Inc  http://www.candelatech.com