Received-SPF: pass (google.com: best guess record for domain of linux-wireless-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
DMARC-Filter: OpenDMARC Filter v1.3.2 smtp.codeaurora.org C926260E7A
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: Re: [PATCH] ath9k: Differentiate between max combined and per chain
 power
From:   Kalle Valo <kvalo@codeaurora.org>
In-Reply-To: <20190320103723.27228-1-sven@narfation.org>
References: <20190320103723.27228-1-sven@narfation.org>
To:     Sven Eckelmann <sven@narfation.org>
Cc:     linux-wireless@vger.kernel.org,
        QCA ath9k Development <ath9k-devel@qca.qualcomm.com>,
        Sven Eckelmann <sven@narfation.org>
User-Agent: pwcli/0.0.0-git (https://github.com/kvalo/pwcli/) Python/2.7.12
Message-Id: <20190429145410.B36C161157@smtp.codeaurora.org>
Date:   Mon, 29 Apr 2019 14:54:10 +0000 (UTC)
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

Sven Eckelmann <sven@narfation.org> wrote:

> The ath9k driver uses as maximum allowed txpower the constant
> MAX_RATE_POWER. It is used to set a maximum txpower limit for the PHY
> (which is combined txpower) and also the maximum txpower for per chain
> rates. Its value 63 is derived from the maximum number the registers can
> store for the per chain txpower.
> 
> The max txpower a user can set because of this is 31 dBm (floor(63 / 2)).
> This also means that a device with multiple tx chains is even limited
> further:
> 
> * 1 chain:  31 dBm per chain
> * 2 chains: 28 dBm per chain
> * 3 chains: 26 dBm per chain
> 
> This combined txpower limit of 31 dBm becomes even more problematic when
> some extra antenna gain is set in the EEPROM. A high power device is then
> no longer able to reach its potential limits.
> 
> Instead the code dealing with the combined txpower must use a higher limit
> than 63 and only the code dealing with the per chain txpower have to use
> the limit of 63. Since the antenna gain can be quite large and 8 bit
> variables are often used in ath9k for txpower, a large, divisible by two
> number like 254 is a good choice for this new limit.
> 
> Signed-off-by: Sven Eckelmann <sven@narfation.org>
> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>

Patch applied to ath-next branch of ath.git, thanks.

b037b107565f ath9k: Differentiate between max combined and per chain power

-- 
https://patchwork.kernel.org/patch/10861389/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches