On Thu, Nov 01, 2012 at 04:28:59PM +0100, Mike Looijmans wrote:
This breaks down if any of the drivers actually uses the reset pin to reset the chip - if the reset might not actually happen but the driver wants it to happen then things are going to go wrong.
Good point. But in that case, the gpio cannot be "optional", the board file MUST supply it.
Well, what normally happens is that the driver has some alternative ways of doing what it wants but if there's a reset signal present it wants to actually use it and be able to rely on it doing the job.
Do I understand correctly that I'm totally alone in this and that there is no point in uploading the code?
If you want to do something like this you should explictly model reset lines like this rather than shoehorning them into a different subsystem.