[U-Boot] preloader_console_init()

[U-Boot] preloader_console_init()

[Simon Schwarz]

Hi Aneesh,

I'am nearly done with OMAP3 - I think I can release the patch today.

I have just one problem left:
You implemented preloader_console_init() in omap-common to call 
setup_clocks_for_console(). This is a OMAP4 specific call - OMAP3 only 
has per_clocks_init() - which inits a bunch of clocks.

Can you change the preloader_console_init()-function that it requires 
that the UART already has a clock and move your 
setup_clocks_for_console() out of the function?

Regards & Thanks
Simon

[U-Boot] preloader_console_init()

[V, Aneesh]

Hi Simon,

On Mon, Jul 25, 2011 at 8:36 PM, Simon Schwarz
<simonschwarzcor@googlemail.com> wrote:
> Hi Aneesh,
>
> I'am nearly done with OMAP3 - I think I can release the patch today.
>
> I have just one problem left:
> You implemented preloader_console_init() in omap-common to call
> setup_clocks_for_console(). This is a OMAP4 specific call - OMAP3 only has
> per_clocks_init() - which inits a bunch of clocks.

Sorry. I was on a short travel and didn't have mail access for a couple of days.

The idea was to enable the console at the earliest, even before prcm_init().
If OMAP3 doesn't need this I would suggest moving setup_clocks_for_console()
call to s_init() in omap4/board.c

BTW, how did you fix this in your patches fianlly?

br,
Aneesh

[U-Boot] preloader_console_init()

[Simon Schwarz]

Hi Aneesh,

On 07/27/2011 04:39 PM, V, Aneesh wrote:
> Hi Simon,
>
> On Mon, Jul 25, 2011 at 8:36 PM, Simon Schwarz
> <simonschwarzcor@googlemail.com>  wrote:
>> Hi Aneesh,
>>
>> I'am nearly done with OMAP3 - I think I can release the patch today.
>>
>> I have just one problem left:
>> You implemented preloader_console_init() in omap-common to call
>> setup_clocks_for_console(). This is a OMAP4 specific call - OMAP3 only has
>> per_clocks_init() - which inits a bunch of clocks.
>
> Sorry. I was on a short travel and didn't have mail access for a couple of days.
>
No problem, hope it was great!

> The idea was to enable the console at the earliest, even before prcm_init().
> If OMAP3 doesn't need this I would suggest moving setup_clocks_for_console()
> call to s_init() in omap4/board.c
So I think it is best to move it to s_init like you said.

For now I don't plan to add this to OMAP3. If we move it to s_init the 
dev can decide to do early clock init or not.
>
> BTW, how did you fix this in your patches fianlly?

#ifndef CONFIG_OMAP34XX
         setup_clocks_for_console();
#endif

To be able to post the patch.

>
> br,
> Aneesh

Regards
Simon