[PATCH] TDMA beacon from slave station

Kang Yin Su cantona at cantona.net
Thu Sep 1 09:56:13 UTC 2011


Hi all,

This is the correction from last TDMA patch.

1. Relocate 'ath_hal_*' functions from if_ath.c to appropriate HAL
layer and moved those common functions to ah.c
2. The next TBTT register mistakenly used in AR6416 that cause
intermittent beacons transmit form a TDMA slave station.
 - Use AR_NEXT_TBTT instead of AR_TIMER0 for AR5416.
3. Function ath_extend_tsf() assume receive time-stamp is in 15bits
TSF that leaks to incorrect beacon timers used for AR5416. The PCU 's
timestamp from AR5416 is 32bits while AR5212 is 15bits.

Thanks,
Yin

On Mon, Aug 29, 2011 at 8:01 PM, Kang Yin Su <cantona at cantona.net> wrote:
> Hi all,
>
> This patch modify the following:
>
> 1. Relocate 'ath_hal_*' functions from if_ath.c to appropriate HAL
> layer and moved those common functions to ah.c
> 2. The next TBTT register mistakenly used in AR6416 that cause
> intermittent beacons transmit form a TDMA slave station.
>  - Use AR_NEXT_TBTT instead of AR_TIMER0 for AR5416.
> 3. Do not adjust the beacon timers when 'tudelta' value is multiple of
> tdmabintval value. It causes the timers to stop and beacon stuck.
>
> Thanks,
> Yin
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 11nnexttbtt_fix_and_hal_func_relocate4.diff
Type: application/octet-stream
Size: 17925 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-wireless/attachments/20110901/247651ea/11nnexttbtt_fix_and_hal_func_relocate4.obj


More information about the freebsd-wireless mailing list