[ath9k-devel] [ath9k_htc] TX retry packet

[ath9k-devel] [ath9k_htc] TX retry packet

[Matteo Danieletto]

Hi all,
I am looking for the number of packets (frame) retransmitted during a
finite interval.
The USB adapter is TL-WN722N (driver ath9k_htc) and I am testing it in IBSS
mode with kernel 3.2.0 .

I found
in /sys/kernel/debug/ieee80211/phy1/netdev:wlan3/stations/00:1d:7d:49:2e:4e/tx_retry_count
but the output is always 0 .

So, for my point of view is strange that never occurred a mac
retransmission.
Do you know if it is a common problem?

I have tried even with an ath9k and this time it worked.

Do you know which is the C struct I need to read to find the number of
retransmission?

Thanks a lot
Best

-- 
Matteo Danieletto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130715/f777bb61/attachment.htm 

[ath9k-devel] [ath9k_htc] TX retry packet

[Oleksij Rempel]

Am 15.07.2013 18:19, schrieb Matteo Danieletto:
> Hi all,
> I am looking for the number of packets (frame) retransmitted during a
> finite interval.
> The USB adapter is TL-WN722N (driver ath9k_htc) and I am testing it in
> IBSS mode with kernel 3.2.0 .
>
> I found
> in /sys/kernel/debug/ieee80211/phy1/netdev:wlan3/stations/00:1d:7d:49:2e:4e/tx_retry_count
>
> but the output is always 0 .
>
> So, for my point of view is strange that never occurred a mac
> retransmission.
> Do you know if it is a common problem?
>
> I have tried even with an ath9k and this time it worked.
>
> Do you know which is the C struct I need to read to find the number of
> retransmission?

Hi Matteo,

you need to digg into firmware. Retransmission is done some where here 
target_firmware/wlan/if_owl.c:
ath_tx_retry_subframe()

or some thing like this. You are welcome to provide patches ;)
see:
https://github.com/qca/open-ath9k-htc-firmware

I also started to work on documentaion... there are some drafts, may be 
it will help you understand how firmware works:

https://github.com/qca/open-ath9k-htc-firmware/wiki/usb-regs

https://docs.google.com/document/d/1pg27qHGx-2hsx9tGCwj0l9EtJZU-viquaOKC02WRt6Y/pub
-- 
Regards,
Oleksij

[ath9k-devel] [ath9k_htc] TX retry packet

[Matteo]

Thanks a lot.
Sorry for delay, but I never received your repley.
Sure, I will try to patch the firmware source code to extract that information.

Thanks again
Matteo

[ath9k-devel] [ath9k_htc] TX retry packet

[Michal Kudelski]

Hi,

Recently, I discovered one problem in firmware, and I 
didn't have time to report it yet... It could also solve 
your issue.

Namely, in the firmware, in the file if_owl.c, in the 
function ath_tx_status_update(...), status flags are not 
being reset. And they definitely should. Thus, you can add 
something like this:

  txs->txstatus[txs->cnt].ts_flags = 0;

after the following line of code:

  txs->txstatus[txs->cnt].ts_rate = SM(bf->bf_endpt, 
ATH9K_HTC_TXSTAT_EPID);

And you need to recompile and re-install the firmware, of 
course.

Hope this can help.

Cheers,
Michal Kudelski


> ------------------------------
> 
> Message: 3
> Date: Mon, 15 Jul 2013 18:19:20 +0200
>From: Matteo Danieletto <matteo.danieletto@gmail.com>
> Subject: [ath9k-devel] [ath9k_htc] TX retry packet
> To: ath9k-devel at lists.ath9k.org, 
>linux-wireless at vger.kernel.org
> Message-ID:
> 	<CAPfemNeQTfPOokxKdvtSk7jYTuT3L+Bf_r=YE5et_v2dVXc9zQ@mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hi all,
> I am looking for the number of packets (frame) 
>retransmitted during a
> finite interval.
> The USB adapter is TL-WN722N (driver ath9k_htc) and I am 
>testing it in IBSS
> mode with kernel 3.2.0 .
> 
> I found
> in 
>/sys/kernel/debug/ieee80211/phy1/netdev:wlan3/stations/00:1d:7d:49:2e:4e/tx_retry_count
> but the output is always 0 .
> 
> So, for my point of view is strange that never occurred 
>a mac
> retransmission.
> Do you know if it is a common problem?
> 
> I have tried even with an ath9k and this time it worked.
> 
> Do you know which is the C struct I need to read to find 
>the number of
> retransmission?
> 
> Thanks a lot
> Best
> 
> -- 
> Matteo Danieletto
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: 
>http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20130715/f777bb61/attachment-0001.htm 
> 
> ------------------------------