Re: [PATCH V4 3/5] arm64: dts: nuvoton: Add initial support for MA35D1

From: Jacky Huang <ychuang570808@gmail.com>
To: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>,
	Jacky Huang <ychuang3@nuvoton.com>,
	linux-kernel@vger.kernel.org, devicetree@vger.kernel.org,
	linux-clk@vger.kernel.org, linux-arm-kernel@lists.infradead.org
Cc: robh+dt@kernel.org, sboyd@kernel.org, krzk+dt@kernel.org,
	arnd@arndb.de, olof@lixom.net, catalin.marinas@arm.com,
	will@kernel.org, soc@kernel.org, cfli0@nuvoton.com
Subject: Re: [PATCH V4 3/5] arm64: dts: nuvoton: Add initial support for MA35D1
Date: Sun, 15 May 2022 13:53:49 +0800	[thread overview]
Message-ID: <db333ce4-6b21-7807-15fa-384619cde6f8@gmail.com> (raw)
In-Reply-To: <0e72e176-d7b0-ed10-08f3-ba7d4729a931@linaro.org>

On 2022/5/13 下午 02:57, Krzysztof Kozlowski wrote:
> On 13/05/2022 08:48, Jacky Huang wrote:
>>>> +
>>>> +	hxt_24m: hxt_24mhz {
>>> No underscores in node name. Generic node names, so "clock-X" or
>>> "clock-some-suffix"
>> OK, I will modify it as
>>    hxt-24m: hxt-24mhz
> No, it is not a generic node name. Please read my reply again.

I  would modify it as

    clock-hxt: clock-hspd-ext-crystal

>
>>>> +		compatible = "fixed-clock";
>>>> +		#clock-cells = <0>;
>>>> +		clock-frequency = <24000000>;
>>> This does not look like property of SoC. Where is this clock defined? In
>>> the SoC or on the board?
>> It's an external crystal on the board.
>> I add this node, because it's the clock source of clock controller.
>> It always present on all ma35d1 boards.
>>
>>       clk: clock-controller@40460200 {
>>           compatible = "nuvoton,ma35d1-clk";
>>           reg = <0x0 0x40460200 0x0 0x100>;
>>           #clock-cells = <1>;
>>           clocks = <&hxt_24m>;
>>           clock-names = "HXT_24MHz";
>> ...
>>
>>>> +		clock-output-names = "HXT_24MHz";
>>>> +	};
>>>> +
>>>> +	timer {
>>>> +		compatible = "arm,armv8-timer";
>>>> +		interrupts = <GIC_PPI 13 (GIC_CPU_MASK_SIMPLE(4) |
>>>> +					  IRQ_TYPE_LEVEL_LOW)>,
>>>> +			     <GIC_PPI 14 (GIC_CPU_MASK_SIMPLE(4) |
>>>> +					  IRQ_TYPE_LEVEL_LOW)>,
>>>> +			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(4) |
>>>> +					  IRQ_TYPE_LEVEL_LOW)>,
>>>> +			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(4) |
>>>> +					  IRQ_TYPE_LEVEL_LOW)>;
>>>> +		clock-frequency = <12000000>;
>>>> +	};
>>>> +
>>>> +	sys: system-controller@40460000 {
>>>> +		compatible = "nuvoton,ma35d1-sys", "syscon", "simple-mfd";
>>> Why is this a simple-mfd if there are no children here? What do you want
>>> to instantiate here?
>> It's not a device, but a set of registers for system level control.
>> I want to provide a register base mapping for other devices to access
>> system control registers.
> This does not explain why you need simple-mfd. simple-mfd is not for
> providing a register base mapping for other devices.

OK, I will remove the "simple-mfd" from sys node.

>>> Where is the nuvoton,ma35d1-sys compatible documented?
>> OK, I will add the compatible document in next version.
>>
>>
>>>> +		reg = <0x0 0x40460000 0x0 0x400>;
>>>> +	};
>>>> +
>>>> +	reset: reset-controller {
>>>> +		compatible = "nuvoton,ma35d1-reset";
>>> Also not documented.
>> I will also add the document for it.
> All of these should fail on checkpatch which points that you either did
> not run it or ignored the result.
>
> Please run checkpatch on all your submissions to Linux kernel and be
> sure that there is no warning or error.
>
>
>
> Best regards,
> Krzysztof

Sure, thank you for reminding.

Sincerely,.

Jacky Huang