[PATCH] staging: tidspbridge: enable watchdog by default

[Ramirez Luna, Omar]

On Thu, Feb 9, 2012 at 9:18 PM, Greg KH <gregkh at linuxfoundation.org> wrote:
> On Thu, Feb 09, 2012 at 04:45:00PM -0800, Ramirez Luna, Omar wrote:
>> Hi,
>>
>> On Thu, Feb 9, 2012 at 3:30 PM, Felipe Contreras
>> <felipe.contreras at gmail.com> wrote:
>> >> Again, I'm totally confused as to _WHY_ this needs to be y.  What is
>> >> causing this oops without it?  If an oops is happening, then shouldn't
>> >> this be a strict dependancy?  Why allow it to be disabled at all if it
>> >> can break your box if you don't enable it?
>> >
>> > It's not an oops, it's a warning, and again, it depends on the
>> > firmware being used. We don't have control over that, and we have no
>> > way to detect if this feature is there. It's up to the user.
>>
>> I have been thinking more into it, how about looking for a WDT symbol
>> inside the baseimage to decide whether to turn ON/OFF WDT3, this would
>> mean that the code is always compiled in, but the decision to turn it
>> on/off is made at runtime.
>
> I totally don't understand, why not just silence the warning properly
> then?
>
> I fail to understand why this warning happens, why it depends on the
> firmware, and why you can't detect it at runtime to not do it.  And how
> it all ties into a kconfig option...

Just FTR, the warning comes from an interconnect driver that reports errors.

This specific warning occurs because the dsp tries to access wdt3
registers without a clock, hence turning ON the wdt3 (or default y for
wdt3 Kconfig) will make the warning to disappear when first loading a
base image into the dsp, because now there will be a clock for the
registers.

It depends on the firmware because the accesses are made from the dsp itself.

Regards,

Omar