On 05/09/2016 05:01 AM, Álvaro Fernández Rojas wrote:

Hello Guenter,

Are there any other issues preventing this patches from being merged?

My major problem/concern is the interrupt handling and the repeated
restart of the hardware timer with continuously reduced timeouts.
This appears fragile, and it doesn't really make sense from a system
level point of view. Unfortunately, I don't have a data sheet for
the chip, nor the means or the time to test it myself. So we are pretty
much in limbo, unless someone from Broadcom confirms that, yes, this is
the one and expected means to program this watchdog. 

Guenter

Regards,
Álvaro.

On 11/28/2015 04:26 AM, Simon Arlott wrote:
 > Add the BCM6345/BCM6318 timer as an interrupt controller so that it can be
 > used by the watchdog to warn that its timer will expire soon.
 >
 > Support for clocksource/clockevents is not implemented as the timer
 > interrupt is not per CPU (except on the BCM6318) and the MIPS clock is
 > better. This could be added later if required without changing the device
 > tree binding.
 >
 > Signed-off-by: Simon Arlott <simon@fire.lp0.eu <mailto:simon@fire.lp0.eu>>

Hi Simon,

can you please re-send the entire series, with all Acked-by:/Reviewed-by:
tags as appropriate ?

Thanks,
Guenter