From mboxrd@z Thu Jan  1 00:00:00 1970
From: nsekhar@ti.com (Sekhar Nori)
Date: Mon, 4 Feb 2013 10:58:46 +0530
Subject: [PATCH v2 6/6] ARM: davinci: da850: configure system configuration
 chip(CFGCHIP3) for emac
In-Reply-To: <510E5BCF.9080006@ti.com>
References: <1359380879-26306-1-git-send-email-prabhakar.lad@ti.com>
 <1359380879-26306-7-git-send-email-prabhakar.lad@ti.com>
 <510E5BCF.9080006@ti.com>
Message-ID: <510F470E.5040205@ti.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On 2/3/2013 6:15 PM, Sekhar Nori wrote:
> On 1/28/2013 7:17 PM, Prabhakar Lad wrote:
>> From: Lad, Prabhakar <prabhakar.lad@ti.com>

> There are couple of issues with this implementation.
> 
> 1) da8xx_config_emac() is specific to DA850 EVM, but masquerades as
>    generic for da8xx. Looks like you need two functions, one for soc
>    specific configuration and one board specific.
> 2) da8xx_config_emac() goes through all the time, whether the
>    particular board has emac module or not. Shouldn't
>    da8xx_config_emac() check if emac is actually enabled in the passed
>    dtb and only the do the configuration?
> 3) The function assumes mii is used always, you can use the rmii_en dt
>    property to check if rmii/mii is enabled and configure the soc/board
>    accordingly.
> 4) If the same function can work both for da850 and da830, then it can
>    be implemented outside of CONFIG_ARCH_DAVINCI_DA850.

Another thing I forgot to mention:

5) This patch replicates a lot of code form board-da850-evm.c. This
needs to be avoided. May be create function to configure emac in
da850.c. This can then be called in board-da850-evm.c as well as in the
dt case.

Thanks,
Sekhar