On Tue 2018-01-02 18:27:20, Sebastian Reichel wrote:

Hi,

On Tue, Jan 02, 2018 at 02:17:22PM +0100, Pavel Machek wrote:

quoted

This adds dts support for magnetometer and touchscreen on Nokia N9.

I think it makes sense to have this splitted.

Creating more work for everyone for little gain? Meh.

quoted

diff --git a/arch/arm/boot/dts/omap3-n9.dts b/arch/arm/boot/dts/omap3-n9.dts
index 39e35f8..57a6679 100644
--- a/arch/arm/boot/dts/omap3-n9.dts
+++ b/arch/arm/boot/dts/omap3-n9.dts

@@ -36,6 +57,22 @@
 			};
 		};
 	};
+
+	touch@4b {

touchscreen@

Ok.

quoted

+		compatible = "atmel,maxtouch";
+		reg = <0x4b>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <29 2>; /* gpio_61, IRQF_TRIGGER_FALLING*/

reset-gpios = <&gpio3 17 GPIO_ACTIVE_SOMETHING>;

quoted

+		vdd-supply = <&vio>;
+		avdd-supply = <&vaux1>;

Those two are not mentioned in the binding and not supported by the
driver as far as I can see?

Driver will need to be fixed, AFAICT :-(.

Touchscreen with the same settings is required for n950, so it
should be in the shared n950 + n9 file.

In future, settings will be different for n9/n950: calibration matrix
is different as panel is rotated in different way. Still it probably
makes sense to share. Ok.

quoted

+&i2c3 {
+	ak8975@0f {
+		compatible = "asahi-kasei,ak8975";
+		reg = <0x0f>;
+	};
 };

Looking at the N9 board file this is missing a rotation matrix. This
is supported by the binding:

Documentation/devicetree/bindings/iio/magnetometer/ak8975.txt

Do you have an idea how the rotation matrix should look like? I don't
currently have an userland software that could calibrate and test the
sensor, so I'd prefer to merge basic binding now and do calibration
later.

Best regards,
									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html