[PATCH 1/3] serial/imx: add device tree support

[Shawn Guo]

On Sat, Jun 18, 2011 at 10:26:55AM -0600, Grant Likely wrote:
> On Sat, Jun 18, 2011 at 06:21:46PM +0200, Arnd Bergmann wrote:
> > On Saturday 18 June 2011 17:19:12 Shawn Guo wrote:
> > >
> > > +Required properties:
> > > +- compatible : should be "fsl,<soc>-uart", "fsl,imx-uart"
> > > +- reg : address and length of the register set for the device
> > > +- interrupts : should contain uart interrupt
> > > +- id : should be the port ID defined by soc
> > > +
> > > +Optional properties:
> > > +- fsl,has-rts-cts : indicate it has rts-cts
> > > +- fsl,irda-mode : support irda mode
> > > +
> > > +Example:
> > > +
> > > +uart at 73fbc000 {
> > > +	compatible = "fsl,imx51-uart", "fsl,imx-uart";
> > > +	reg = <0x73fbc000 0x4000>;
> > > +	interrupts = <31>;
> > > +	id = <1>;
> > > +	fsl,has-rts-cts;
> > > +};
> > 
> > Should this also support the "clock-frequency" property that 8250-style
> > serial ports support [1]?
> > 
> > For the flow-control properties, should we name that more generic? The
> > same property certainly makes sense for other serial-ports if it does
> > here. OTOH, I'm not sure it's actually reliable, because it also depends
> > on whether the other side of the connection and the cable support hw flow
> > control.
> 
> I'd like to see a few use cases before defining a common property.
> That said, has-rts-cts does sound like a useful generic property.
> Or maybe named "uart-has-rts-cts" to make it clear that it is a uart
> specific binding?
> 
I would keep the name as it is if you do not mind, since it's been
under uart node.

-- 
Regards,
Shawn