Re: [PATCH v5 7/7] arm64: dts: mediatek: Initial mt8365-evk support

[AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>]

Il 14/12/22 00:43, Bernhard Rosenkränzer ha scritto:
> From: Fabien Parent <fparent@baylibre.com>
> 
> This adds minimal support for the Mediatek 8365 SOC and the EVK reference
> board, allowing the board to boot to initramfs with serial port I/O.
> 
> Signed-off-by: Fabien Parent <fparent@baylibre.com>
> [bero@baylibre.com: Removed parts depending on drivers that aren't upstream yet, cleanups, add L2 cache]
> Signed-off-by: Bernhard Rosenkränzer <bero@baylibre.com>
> Tested-by: Kevin Hilman <khilman@baylibre.com>
> ---
>   arch/arm64/boot/dts/mediatek/Makefile       |   1 +
>   arch/arm64/boot/dts/mediatek/mt8365-evk.dts | 163 ++++++++++
>   arch/arm64/boot/dts/mediatek/mt8365.dtsi    | 343 ++++++++++++++++++++
>   3 files changed, 507 insertions(+)
>   create mode 100644 arch/arm64/boot/dts/mediatek/mt8365-evk.dts
>   create mode 100644 arch/arm64/boot/dts/mediatek/mt8365.dtsi
> 

..snip..

> diff --git a/arch/arm64/boot/dts/mediatek/mt8365.dtsi b/arch/arm64/boot/dts/mediatek/mt8365.dtsi
> new file mode 100644
> index 0000000000000..2c4ef9b92b68b
> --- /dev/null
> +++ b/arch/arm64/boot/dts/mediatek/mt8365.dtsi
> @@ -0,0 +1,343 @@
> +// SPDX-License-Identifier: (GPL-2.0 OR MIT)
> +/*
> + * (C) 2018 MediaTek Inc.
> + * Copyright (C) 2022 BayLibre SAS
> + * Fabien Parent <fparent@baylibre.com>
> + * Bernhard Rosenkränzer <bero@baylibre.com>
> + */
> +#include <dt-bindings/clock/mediatek,mt8365-clk.h>
> +#include <dt-bindings/interrupt-controller/arm-gic.h>
> +#include <dt-bindings/interrupt-controller/irq.h>
> +#include <dt-bindings/phy/phy.h>
> +#include <dt-bindings/thermal/thermal.h>
> +
> +/ {
> +	compatible = "mediatek,mt8365";
> +	interrupt-parent = <&sysirq>;
> +	#address-cells = <2>;
> +	#size-cells = <2>;
> +
> +	cpus: cpus {

You're not referencing `cpus` anywhere, hence this label is useless:
please remove.

> +		#address-cells = <1>;
> +		#size-cells = <0>;
> +
> +		cpu-map {
> +			cluster0: cluster0 {

Same for this one.

> +				core0 {
> +					cpu = <&cpu0>;
> +				};
> +				core1 {
> +					cpu = <&cpu1>;
> +				};
> +				core2 {
> +					cpu = <&cpu2>;
> +				};
> +				core3 {
> +					cpu = <&cpu3>;
> +				};
> +			};
> +		};
> +
> +		cpu0: cpu@0 {
> +			device_type = "cpu";
> +			compatible = "arm,cortex-a53";
> +			reg = <0x0>;
> +			#cooling-cells = <2>;
> +			enable-method = "psci";
> +			next-level-cache = <&l2>;

It would be nice if you described the I/D caches for all CPUs.

> +		};
> +
> +		cpu1: cpu@1 {
> +			device_type = "cpu";
> +			compatible = "arm,cortex-a53";
> +			reg = <0x1>;
> +			#cooling-cells = <2>;
> +			enable-method = "psci";
> +			next-level-cache = <&l2>;
> +		};
> +
> +		cpu2: cpu@2 {
> +			device_type = "cpu";
> +			compatible = "arm,cortex-a53";
> +			reg = <0x2>;
> +			#cooling-cells = <2>;
> +			enable-method = "psci";
> +			next-level-cache = <&l2>;
> +		};
> +
> +		cpu3: cpu@3 {
> +			device_type = "cpu";
> +			compatible = "arm,cortex-a53";
> +			reg = <0x3>;
> +			#cooling-cells = <2>;
> +			enable-method = "psci";
> +			next-level-cache = <&l2>;
> +		};
> +
> +		l2: l2-cache {
> +			compatible = "cache";

....and what's the size of this L2 cache?
Is it unified, or does each CPU core have its own private L2?

> +		};
> +	};
> +
> +	clk26m: oscillator {
> +		compatible = "fixed-clock";
> +		#clock-cells = <0>;
> +		clock-frequency = <26000000>;
> +		clock-output-names = "clk26m";
> +	};
> +
> +	psci {
> +		compatible = "arm,psci-1.0";
> +		method = "smc";
> +	};
> +
> +	reserved-memory {
> +		#address-cells = <2>;
> +		#size-cells = <2>;
> +		ranges;
> +
> +		/* 128 KiB reserved for ARM Trusted Firmware (BL31) */
> +		bl31_secmon_reserved: secmon@43000000 {

This depends on the bootloader that's flashed on your board - it's not a
global SoC property.

Please move it to your board, or explain why BL31 *must* always be
128KiB starting at 0x43000000.

> +			no-map;
> +			reg = <0 0x43000000 0 0x20000>;
> +		};
> +	};
> +
> +	soc {
> +		#address-cells = <2>;
> +		#size-cells = <2>;
> +		compatible = "simple-bus";
> +		ranges;
> +

Is there really no systimer in this SoC? Would be pretty odd....

Regards,
Angelo