MIME-Version: 1.0
In-Reply-To: <AANLkTi=YLPnqRzzfXaeghQ-w5vDgfCX5v59tGbgSp8zA@mail.gmail.com>
References: <20101123191533.GP4303@makis.mantri>
	<AANLkTi=YLPnqRzzfXaeghQ-w5vDgfCX5v59tGbgSp8zA@mail.gmail.com>
Date: Wed, 24 Nov 2010 14:55:54 +0200
Message-ID: <AANLkTimEZnDrXYa4WOKn=9VvL0qwYUTbABy1WgmSqe4m@mail.gmail.com>
Subject: Re: [ath5k-devel] [PATCH 16/30] ath5k: Set all IFS intervals, not
 just slot time
From: Nick Kossifidis <mickflemm@gmail.com>
To: Jonathan Guerin <jonathan@guerin.id.au>
Cc: ath5k-devel@lists.ath5k.org, linux-wireless@vger.kernel.org,
	linville@tuxdriver.com, me@bobcopeland.com, mcgrof@gmail.com,
	jirislaby@gmail.com, nbd@openwrt.org, br1@einfach.org
Content-Type: text/plain; charset=UTF-8
Sender: linux-wireless-owner@vger.kernel.org

2010/11/24 Jonathan Guerin <jonathan@guerin.id.au>:
>
>> diff --git a/drivers/net/wireless/ath/ath5k/pcu.c b/drivers/net/wireless/ath/ath5k/pcu.c
>> index e691378..4556f29 100644
>> --- a/drivers/net/wireless/ath/ath5k/pcu.c
>> +++ b/drivers/net/wireless/ath/ath5k/pcu.c
>> @@ -763,7 +763,7 @@ ath5k_hw_check_beacon_timers(struct ath5k_hw *ah, int intval)
>>  * @ah: The &struct ath5k_hw
>>  * @coverage_class: IEEE 802.11 coverage class number
>>  *
>> - * Sets slot time, ACK timeout and CTS timeout for given coverage class.
>> + * Sets IFS intervals and ACK/CTS timeouts for given coverage class.
>>  */
>>  void ath5k_hw_set_coverage_class(struct ath5k_hw *ah, u8 coverage_class)
>>  {
>> @@ -772,7 +772,7 @@ void ath5k_hw_set_coverage_class(struct ath5k_hw *ah, u8 coverage_class)
>>        int ack_timeout = ath5k_hw_get_default_sifs(ah) + slot_time;
>
> This is not quite right:
>
> According to the 802.11-2007 spec document, the ACKTimeout value is
> (Section 9.2.8 ACK procedure):
> ACKTimeout = aSIFSTime + aSlotTime + aPHY-RX-START-Delay
>
> From Table 17-15—OFDM PHY characteristics, the values are:
> aSIFSTime = 16
> aSlotTime = 9
> aPHY-RX-START-Delay = 25
>
> Therefore, ACKTimeout = 50
>
> Ignoring my uniformed comments from before, this is the only thing I
> can see that's wrong.
>
> Cheers,
>
> Jonathan

Hmm I didn't mess with set_coverage_class so i didn't look up for ack
timeout. That phy-rx-start-delay is standard value or hw specific ?
Also does it change with clockrate (bwmodes) ? We already have a phy
activation -> rx start delay (check patch 25). It's 10.000 on RF5111
and 2.000 on RF5112 and later, if we divide by A clock it's 250 on
RF5111 and 50 on RF5112 and later. Do you think it's related ?

-- 
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick