"Mark A. Greer" <[email protected]> writes:

> On Mon, Mar 30, 2009 at 12:53:29PM -0700, Kevin Hilman wrote:
>> "Mark A. Greer" <[email protected]> writes:
>> 
>> > From: Mark A. Greer <[email protected]>
>> >
>> > The watchdog code currently hardcodes the base address
>> > of the timer its using.  To support new SoCs, make it
>> > support timers at any address.  Use davinci_soc_info
>> > to do this.
>> >
>> > Signed-off-by: Mark A. Greer <[email protected]>
>> 
>> You should also update the WDT setup from devices.c.  This is where
>> it sets up the actual watchdog driver.
>> 
>> Now that it's in soc_info, it could be done in common.c instead of devices.c.
>
> You mean doing the platform_device_register(), right?  If so, I
> think I'll put it in the soc files then

I don't necessarily care about moving the platform_device_register() per-se.

What I don't like is the fact that the watchdog bases are hard-coded
in devices.c.  Now that they are defined in SoC specific code, that
devices.c code could get the base from soc_info instead of hard-coding
it.

> because the da830 stuff does the registering a bit differently (see
> devices-da830.c:da830_add_devices()).

And I don't like (or maybe don't get) that registration method very
much either.  It seems like a needless and confusing layer of
redirection.  Or maybe it's because I have only looked at it when it's
been past my bedtime, so my chances of understanding it are slim. ;)

Kevin


_______________________________________________
Davinci-linux-open-source mailing list
[email protected]
http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source

Reply via email to