Re: AR9287 ; 2-wire coexistence expected behavior

[Adrian Chadd <adrian@freebsd.org>]

You'll have to give me some more time to digest what I'm reading
internally; it's all new to me.

For FreeBSD, you can look at these files in FreeBSD-HEAD:

sys/dev/ath/ath_hal/ar5416/ar5416_btcoex.c
sys/dev/ath/ath_hal/ar9002/ar9285_btcoex.c (not that relevant for you,
but still)

These contain the register writes that the internal reference HAL is
doing when programming btcoex for AR9285/AR9287.

The AR9287 should support 2-wire coex fine.

Just read those source files. The ar5416InitBTCoex() programs the
BT_ACTIVE / WLAN_ACTIVE gpio's when you say the coexistence type is
2-wire. There's weights and such being programmed elsewhere; you need
to make sure you pick the default values for that and try it out.

I can't help you more than that at the moment; I'm busy doing other
work that doesn't at all touch wifi. :)



Adrian

On 4 April 2013 20:08, sandeep suresh <sandeep.suresh@yahoo.co.in> wrote:
> Hello Mr.Adrian,
>     Thanks for your mail. Regarding your comment below:
>
> "THen you need to ensure the bluetooth coexistence registers are programmed
> so that btactive will correctly stomp wifi traffic."
>
> Can you please elaborate on this? As I understand the concept of "stomping"
> is only in 3-wire coexistence; correct me if I am wrong.
>
> As mentioned earlier, the set-up we have is a PCB board with general purpose
> MCU controlling the 2-wire coexistence pins of AR9287. As there are other
> 2.4GHz radio (called PROP radio here) on the board, we want to ensure that
> both radios do not transmit at the same time which will result in
> collisions. Hence we want to build a co-operative coexistence approach so
> that when PROP radio is active (by asserting BT_ACTIVE), AR9287 has to
> buffer transmissions and allow PROP radio to be active. Only when PROP radio
> is inactive (BT_ACTIVE is deasserted), only then WiFi can be active. Hope
> this is achievable?
>
> Thanks & Regards
> Sandeep.
>
> From: Adrian Chadd <adrian@freebsd.org>
> To: sandeep suresh <sandeep.suresh@yahoo.co.in>
> Cc: "linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>;
> ath9k-devel <ath9k-devel@lists.ath9k.org>
> Sent: Thursday, 4 April 2013 11:36 PM
>
> Subject: Re: AR9287 ; 2-wire coexistence expected behavior
>
> Hi!
>
> I'm glad you're looking into this in more depth!
>
> On 4 April 2013 08:19, sandeep suresh <sandeep.suresh@yahoo.co.in> wrote:
>
>> I understand that ATH_BTCOEX_CFG_2_WIRE, ATH_BTACTIVE_GPIO_9280 (GPIO6 as
>> per btcoex.h) and ATH_WLANACTIVE_GPIO_9280 (GPIO5 as per btcoex.h) are
>> used.
>
> Ok, right.
>
>> I next started monitoring GPIO5 on oscillaoscope to see WLAN activity and
>> I
>> could see a lot of pulse trains. Next in order to simulate high priority
>> BT
>> traffic, I pulled the line GPIO6 high. But I did not see any change in
>> WLAN
>> activity as I could continue to see the pulse trains. My expectation was
>> that there should not be any WLAN activity and hence no pulses. Please
>> guide
>> if I am missing anything?
>
> Well, firstyl you need to ensure that the GPIO pin has been programmed
> to be an input, and it's of the right bluetooth type. As I said
> before, GPIO pins can be input, output; they can be connected via an
> internal mux to a variety of "behaviours". Take a look at the gpio
> configure code in ath9k to see more.
>
> THen you need to ensure the bluetooth coexistence registers are
> programmed so that btactive will correctly stomp wifi traffic.
>
> THat's all I know for now. Sorry.
>
>
>
> adrian
>
>