On Mon, 17 Jul 2017 11:41:01 +0530
Abhishek Sahu [off-list ref] wrote:

quoted

quoted

quoted

+
+nand@79b0000 {  

nand-controller@xxxx {

BTW, glad to see another driver moving to the new DT representation 
:-).
  

quoted

quoted

+	compatible = "qcom,qpic-nandc-v1.4.0";
+	reg = <0x79b0000 0x1000>;
+
+	clocks = <&gcc GCC_QPIC_CLK>,
+		<&gcc GCC_QPIC_AHB_CLK>;
+	clock-names = "core", "aon";
+
+	#address-cells = <1>;
+	#size-cells = <0>;
+
+	nandcs@0 {  

	nand@0 {
  

quoted

quoted

+		compatible = "qcom,nandcs";  

Why do you need a compatible here?  

It is the part of original driver. We can connect multiple
NAND devices in the same bus and qcom,nandcs is being used
for each connected NAND device. Each NAND device can use
different  chip select, ecc strength etc which we can specify
under this sub node.

Still don't understand why you need a compatible? Is this a memory bus
where you can connect other kind of memories (parallel NORs,
SRAMs, ...)?

If that's not the case, then considering all subnodes of the
nand-controller node containing a reg property as NAND devices is fine,
you don't need this compatible = "nand,cs" (see sunxi-nand bindings
[1]).

If the bus is generic and can be attached non-NAND devices, I'd
recommend looking at atmel's binding [2], because you're likely to
have one instance of the NAND controller logic for all NAND devices
connected on this bus.
And more importantly, if the bus a generic, the node should not be
named nand or nand-controller, and the compatible should not contain
'nandc' in it.

[1]http://elixir.free-electrons.com/linux/latest/source/Documentation/devicetree/bindings/mtd/sunxi-nand.txt#L34
[2]http://elixir.free-electrons.com/linux/latest/source/Documentation/devicetree/bindings/mtd/atmel-nand.txt#L70
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html