From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Date: Thu, 30 Apr 2009 01:19:06 +0200
Subject: [U-Boot] [PATCH 3/3][Net][ARM] Switched dm644x_emac
	Ethernet	driver to use newer API
In-Reply-To: <49F8D0A1.6010407@gmail.com>
References: <1240938733-3121-1-git-send-email-biggerbadderben@gmail.com>
	<1240938733-3121-2-git-send-email-biggerbadderben@gmail.com>
	<1240938733-3121-3-git-send-email-biggerbadderben@gmail.com>
	<1240938733-3121-4-git-send-email-biggerbadderben@gmail.com>
	<20090429201335.GB522@game.jcrosoft.org>
	<49F8B8B5.9070703@gmail.com>
	<20090429204937.GC522@game.jcrosoft.org>
	<49F8D0A1.6010407@gmail.com>
Message-ID: <20090429231906.GM522@game.jcrosoft.org>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

On 15:11 Wed 29 Apr     , Ben Warren wrote:
> Jean-Christophe PLAGNIOL-VILLARD wrote:
>> On 13:29 Wed 29 Apr     , Ben Warren wrote:
>>   
>>> Hi J-C,
>>>
>>> Jean-Christophe PLAGNIOL-VILLARD wrote:
>>> <snip>
>>>     
>>>>> +/*
>>>>> + *  * Initializes on-chip ethernet controllers.
>>>>> + *   * to override, implement board_eth_init()
>>>>> + *    */
>>>>> +int cpu_eth_init(bd_t *bis)
>>>>> +{
>>>>> +#if defined(CONFIG_DRIVER_TI_EMAC)
>>>>> +	dm644x_emac_initialize();
>>>>> +#endif
>>>>> +	return 0;
>>>>> +}
>>>>>             
>>>> please move this to the soc
>>>> arm926ejs/davinci/cpu.c
>>>>
>>>> the cpu.c is destinated to be removed in most of the case
>>>> or at least all the soc specific code
>>>>
>>>> Best Regards,
>>>> J.
>>>>         
>>> There is currently no such file (not even in arm/next), or that's 
>>> where  this would have gone.  This function needs to be in a source 
>>> file that  already has strongly-linked symbols or it won't override 
>>> the weak  version in net/eth.c.  I'm certainly open to suggestions.
>>>     
>> or simply add an entry in the lds this will force the file to be evaluated
>> first
>>
>>   
> That's not a scalable solution and isn't how we did it with other  
> controllers.  If there's no file that's guaranteed to be included in all  
> Davinci boards that will have this MAC, an alternative is to implement  
> board_eth_init() on each board, but IMHO that's worse architecturally.
maybe sync with David

as the clock function could be move to cpu.c it will have the stringly-link
function

Best Regards,
J.