[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Oleksij Rempel]

Hallo all,

i was playing with STBC functionality and wont to see more statistic 
about it. If i'm correct Some MACs can report if it got STBC frame or at 
least report duplicat frames. After greping the source, i'm confused 
about location of this registers. If i'm correct, STBC and Duplicate 
status bits loctaed together with GI and 2040. According to the source:

ar9003_mac.c:	rxs->rs_flags  = (rxsp->status4 & AR_GI) ? ATH9K_RX_GI : 0;
mac.c:		(ads.ds_rxstatus3 & AR_GI) ? ATH9K_RX_GI : 0;

So i it true for ar9280/ar9285/ar9287 and too? Shouldn't that be 
ds_rxstatus7 or some thing like this? And if it is different where 
should i add statistic for this chips?
-- 
Regards,
Oleksij

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Adrian Chadd]

Hi,

There's no way to know that you received an STBC encoded frame. :(

GI is short/long guard interval, not STBC.



Adrian

On 27 April 2013 01:08, Oleksij Rempel <linux@rempel-privat.de> wrote:
> Hallo all,
>
> i was playing with STBC functionality and wont to see more statistic
> about it. If i'm correct Some MACs can report if it got STBC frame or at
> least report duplicat frames. After greping the source, i'm confused
> about location of this registers. If i'm correct, STBC and Duplicate
> status bits loctaed together with GI and 2040. According to the source:
>
> ar9003_mac.c:   rxs->rs_flags  = (rxsp->status4 & AR_GI) ? ATH9K_RX_GI : 0;
> mac.c:          (ads.ds_rxstatus3 & AR_GI) ? ATH9K_RX_GI : 0;
>
> So i it true for ar9280/ar9285/ar9287 and too? Shouldn't that be
> ds_rxstatus7 or some thing like this? And if it is different where
> should i add statistic for this chips?
> --
> Regards,
> Oleksij
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Oleksij Rempel]

Am 27.04.2013 10:27, schrieb Adrian Chadd:
> Hi,
>
> There's no way to know that you received an STBC encoded frame. :(
>
> GI is short/long guard interval, not STBC.


I do not mean encoded frame, i mean only status flag. Looks like ar9271 
provide it.

>
>
> Adrian
>
> On 27 April 2013 01:08, Oleksij Rempel <linux@rempel-privat.de> wrote:
>> Hallo all,
>>
>> i was playing with STBC functionality and wont to see more statistic
>> about it. If i'm correct Some MACs can report if it got STBC frame or at
>> least report duplicat frames. After greping the source, i'm confused
>> about location of this registers. If i'm correct, STBC and Duplicate
>> status bits loctaed together with GI and 2040. According to the source:
>>
>> ar9003_mac.c:   rxs->rs_flags  = (rxsp->status4 & AR_GI) ? ATH9K_RX_GI : 0;
>> mac.c:          (ads.ds_rxstatus3 & AR_GI) ? ATH9K_RX_GI : 0;
>>
>> So i it true for ar9280/ar9285/ar9287 and too? Shouldn't that be
>> ds_rxstatus7 or some thing like this? And if it is different where
>> should i add statistic for this chips?
>> --
>> Regards,
>> Oleksij
>> _______________________________________________
>> ath9k-devel mailing list
>> ath9k-devel at lists.ath9k.org
>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel


-- 
Regards,
Oleksij

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Adrian Chadd]

On 27 April 2013 01:30, Oleksij Rempel <linux@rempel-privat.de> wrote:

> I do not mean encoded frame, i mean only status flag. Looks like ar9271
> provide it.
>

There's an "STBC received" status flag in the RX decriptor? Where?



adrian

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Oleksij Rempel]

Am 27.04.2013 10:30, schrieb Oleksij Rempel:
> Am 27.04.2013 10:27, schrieb Adrian Chadd:
>> Hi,
>>
>> There's no way to know that you received an STBC encoded frame. :(
>>
>> GI is short/long guard interval, not STBC.

i know, but if i do not have documentation, i combine parts of what i 
know and what i find. I know rx_GI bit is next to 2040 and Duplicate 
bit. Is it correct?

>
> I do not mean encoded frame, i mean only status flag. Looks like ar9271
> provide it.

oops, i misread "to know"... but i still mean the same. Beside, 
Duplicate may appear on STBC too. IS it correct?

-- 
Regards,
Oleksij

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Oleksij Rempel]

Am 27.04.2013 10:53, schrieb Adrian Chadd:
> On 27 April 2013 01:30, Oleksij Rempel <linux@rempel-privat.de> wrote:
>
>> I do not mean encoded frame, i mean only status flag. Looks like ar9271
>> provide it.
>>
>
> There's an "STBC received" status flag in the RX decriptor? Where?

7. Word, 3. bit.

-- 
Regards,
Oleksij

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Adrian Chadd]

On 27 April 2013 01:55, Oleksij Rempel <linux@rempel-privat.de> wrote:

>> There's an "STBC received" status flag in the RX decriptor? Where?
>
> 7. Word, 3. bit.

Where'd you source that from?

The datasheets don't say that. :)



Adrian

[ath9k-devel] rx status registers for ar9280/ar9825/ar9287

[Adrian Chadd]

On 27 April 2013 09:00, Adrian Chadd <adrian@freebsd.org> wrote:
> On 27 April 2013 01:55, Oleksij Rempel <linux@rempel-privat.de> wrote:
>
>>> There's an "STBC received" status flag in the RX decriptor? Where?
>>
>> 7. Word, 3. bit.
>
> Where'd you source that from?
>
> The datasheets don't say that. :)

Hey look! The datasheets are wrong. Shocking!

I went digging through some internal stuff and I found that you're
actually right, and right for the AR9280 and later chips.

For the AR9280 and later:


3       3       stbc                    Receive packet STBC indicator.
 If this value is set then the
                                        baseband has received an STBC
frames as indicated in the HT_PLCP.

So, if you want to patch the pre-AR9380 path in mac80211, you should
check that the chip is at least an AR9280. AR9285 and AR9287 is fine.
AR5416/AR9130/AR9160 isn't - those chips don't support STBC RX.

Grr, another thing to add to FreeBSD today. :-)


Adrian