Re: [PATCH] mtd: gpio-nand: add device tree bindings

[Jamie Iles <jamie@jamieiles.com>]

Hmm, get_maintainer.pl got Artem's address wrong (old Nokia address) for 
gpio-nand.c so correct one on CC now!

On Wed, Jul 27, 2011 at 02:45:01PM -0500, Scott Wood wrote:
> On Wed, 27 Jul 2011 15:03:30 +0100
> Jamie Iles <jamie@jamieiles.com> wrote:
> 
> > diff --git a/Documentation/devicetree/bindings/mtd/gpio-nand.txt b/Documentation/devicetree/bindings/mtd/gpio-nand.txt
> > new file mode 100644
> > index 0000000..98cb152
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/mtd/gpio-nand.txt
> > @@ -0,0 +1,43 @@
> > +GPIO assisted NAND flash
> > +
> > +Required properties:
> > +- compatible : "gpio-nand"
> > +- reg : should specify localbus chip select and size used for the chip.  For
> > +  ARM platforms where a dummy read is needed to provide synchronisation with
> > +  regards to bus reordering, an optional second resource describes the
> > +  location to read from.
> 
> I don't see how a pure "gpio nand" device would have any memory mapped
> I/O.  I think you need a more specific compatible for this.

OK, fair point.  I'm not sure what a better name would be though, maybe 
gpio-assisted-nand?

> > +Optional properties:
> > +- bank-width : Width (in bytes) of the bank.  Equal to the device width times
> > +  the number of interleaved chips.
> 
> Interleaved NAND chips?  Is that actually done?

Doh, that shouldn't read like that.  It's really just the bank width.

> > +Examples:
> > +
> > +gpio-nand@1,0 {
> > +	compatible = "gpio-nand";
> > +	reg = <1 0x0000 0x1000>;
> > +	#address-cells = <1>;
> > +	#size-cells = <1>;
> > +	gpios = <&banka 1 0	/* rdy */
> > +		 &banka 2 0 	/* nce */
> > +		 &banka 3 0 	/* ale */
> > +		 &banka 4 0 	/* cle */
> > +		 0		/* nwp */>;
> > +
> > +	flash {
> > +		#address-cells = <1>;
> > +		#size-cells = <1>;
> > +		compatible = "...";
> > +
> > +		partition@0 {
> > +			...
> > +		};
> > +	};
> > +};
> 
> Here you have a separate flash node underneath the gpio-nand node, but
> earlier in the patch comment you show the partitions being directly under
> gpio-nand, and from a quick glance it appears the latter is what the code
> supports.

Yes, that's definitely wrong!  The partitions should be directly under the 
nand controller node.

Thanks for the review!

Jamie