Hi Neil

see my comments in line

On 11/17/17 21:05, Neil Armstrong wrote:

Hi Yixun, Jian,

On 17/11/2017 09:02, Yixun Lan wrote:

quoted

From: Jian Hu <redacted>

There are four I2C masters in EE domain, and one I2C Master in
AO domain, the DT info here should describe them all.

Signed-off-by: Jian Hu <redacted>
Signed-off-by: Yixun Lan <redacted>
---
 arch/arm64/boot/dts/amlogic/meson-axg.dtsi | 59 ++++++++++++++++++++++++++++++
 1 file changed, 59 insertions(+)

diff --git a/arch/arm64/boot/dts/amlogic/meson-axg.dtsi b/arch/arm64/boot/dts/amlogic/meson-axg.dtsi
index 57faaa9d8013..99e967aff439 100644
--- a/arch/arm64/boot/dts/amlogic/meson-axg.dtsi
+++ b/arch/arm64/boot/dts/amlogic/meson-axg.dtsi

@@ -129,6 +129,54 @@
 				#reset-cells = <1>;
 			};
 
+			i2c_m0: i2c at 1f000 {

Can you use a more simple "i2c0" phandle naming instead ? or get back to the "i2c_A/B/C/D" naming like in meson_gx to keep consistency ?

Ok, I would prefer to use 'i2c0' then

we had a internal discussion about the inconsistent naming of the I2C
controller, and reach a consensus to name them as:
 the EE domain: ee_i2c_m0, ee_i2c_m1 ..
 the AO domain: ao_i2c_m0, ao_i2c_s0
in the future datasheet.

note: 'ee' means the EE domain, m0 mean the master 0, s0 mean the I2C
slave 0

[...]

quoted

+			};
+
+			i2c_m1: i2c at 1e000 {

[...]

quoted

+			};
+
+			i2c_m2: i2c at 1d000 {

[...]

quoted

+			};
+
+			i2c_m3: i2c at 1c000 {

[...]

quoted

+			};
+
 			uart_A: serial at 24000 {
 				compatible = "amlogic,meson-gx-uart", "amlogic,meson-uart";
 				reg = <0x0 0x24000 0x0 0x14>;

@@ -312,6 +360,17 @@
 				};
 			};
 
+			i2c_ao: i2c at 5000 {

Can you keep the same "i2c_AO" as in meson-gx ?

sure, can do

[...]

quoted

+			};
+
 			uart_AO: serial at 3000 {
 				compatible = "amlogic,meson-gx-uart", "amlogic,meson-ao-uart";
 				reg = <0x0 0x3000 0x0 0x18>;

Neil

.