Re: [PATCH v4 2/2] dt-bindings: nand: meson: refine Amlogic NAND controller driver

From: Miquel Raynal <miquel.raynal@bootlin.com>
To: Liang Yang <liang.yang@amlogic.com>
Cc: <linux-mtd@lists.infradead.org>, Rob Herring <robh+dt@kernel.org>,
	Richard Weinberger <richard@nod.at>,
	Vignesh Raghavendra <vigneshr@ti.com>,
	Jerome Brunet <jbrunet@baylibre.com>,
	Neil Armstrong <narmstrong@baylibre.com>,
	Martin Blumenstingl <martin.blumenstingl@googlemail.com>,
	Kevin Hilman <khilman@baylibre.com>,
	Jianxin Pan <jianxin.pan@amlogic.com>,
	Victor Wan <victor.wan@amlogic.com>,
	XianWei Zhao <xianwei.zhao@amlogic.com>,
	Kelvin Zhang <kelvin.zhang@amlogic.com>,
	BiChao Zheng <bichao.zheng@amlogic.com>,
	YongHui Yu <yonghui.yu@amlogic.com>,
	<linux-arm-kernel@lists.infradead.org>,
	<linux-amlogic@lists.infradead.org>,
	<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v4 2/2] dt-bindings: nand: meson: refine Amlogic NAND controller driver
Date: Wed, 20 Apr 2022 14:16:20 +0200	[thread overview]
Message-ID: <20220420141620.4fd68eef@xps13> (raw)
In-Reply-To: <b880c64c-7651-c445-4e5e-74cb7a1e76ee@amlogic.com>

Hi Liang,

> >> +maintainers:
> >> +  - liang.yang@amlogic.com
> >> +
> >> +properties:
> >> +  compatible:
> >> +    enum:
> >> +      - "amlogic,meson-gxl-nfc"
> >> +      - "amlogic,meson-axg-nfc"
> >> +
> >> +  reg:
> >> +    maxItems: 2
> >> +
> >> +  '#address-cells':
> >> +    const: 1  
> > 
> > Not sure this property is needed.  
> this is for the subnode, such as nand@0.

Yes but if you refer to nand-controller.yaml you no longer need these.

> >   
> >> +
> >> +  '#size-cells':
> >> +    const: 0  
> > 
> > Ditto. Plus, this one looks wrong anyway.  
> this is for the subnode, such as nand@0. do you mean s/''/""/?

Sorry, this is not "wrong anyway", my fault. But still, you don't need
this property for the same reason as above.

> >   
> >> +
> >> +  reg-names:
> >> +    items:
> >> +      - const: nfc
> >> +      - const: emmc  
> > 
> > Why do you need the emmc register map? Do you really need to perform a
> > register access there?  
> yes, we have to access the emmc register map. because the NFC clock comes from SDEMMC_CLOCK register.

But if it's a clock you should get the clock and call
clk_prepare_enable(), you don't need to poke directly in the registers.
Do you?

> >> +examples:
> >> +  - |
> >> +    #include <dt-bindings/clock/axg-clkc.h>
> >> +    #include <dt-bindings/interrupt-controller/arm-gic.h>
> >> +    apb {
> >> +      #address-cells = <2>;
> >> +      #size-cells = <2>;  
> > 
> > Not sure you need this upper node in the example.  
> use the upper node to indicate the "#address-cells" and "#size-cells". if i do not do that, dt_binding_check will report:
>   ".....reg:0: [0, 30720, 0, 256] is too long" and
>   ".....reg:1: [0, 28672, 0, 2048] is too long".

ok, maybe, I'll let bindings maintainer review that.

> >   
> >> +      nand-controller@7800 {
> >> +        #address-cells = <1>;
> >> +        #size-cells = <0>;
> >> +        compatible = "amlogic,meson-axg-nfc";
> >> +        reg = <0x0 0x7800 0x0 0x100>,
> >> +              <0x0 0x7000 0x0 0x800>;
> >> +        reg-names = "nfc", "emmc";
> >> +
> >> +        interrupts = <GIC_SPI 34 IRQ_TYPE_EDGE_RISING>;
> >> +        clocks = <&clkc CLKID_SD_EMMC_C>,
> >> +                 <&clkc CLKID_FCLK_DIV2>;
> >> +        clock-names = "core", "device";
> >> +
> >> +      };
> >> +    };
> >> +...  

Thanks,
Miquèl

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel