Re: [PATCH v3 1/3] phy: Add exynos-simple-phy driver

[Thierry Reding <thierry.reding@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2447 bytes --]

On Thu, May 15, 2014 at 12:47:21AM +0530, Rahul Sharma wrote:
[...]
> diff --git a/Documentation/devicetree/bindings/phy/samsung-phy.txt b/Documentation/devicetree/bindings/phy/samsung-phy.txt
[...]
> +For "samsung,exynos4210-simple-phy" compatible PHYs the second cell in
> +the PHY specifier identifies the PHY and the supported phys for exynos4210

I think the specifier is only the part after the phandle, so this should
probably be "... compatible PHYs the single cell specifier ..." or
something equivalent.

> +are:
> +  HDMI_PHY,
> +  DAC_PHY,
> +  ADC_PHY,
> +  PCIE_PHY,
> +  SATA_PHY.

I think you need to specify the literal values here as well, since the
binding must be fully self-contained. That is you can't rely on the DT
binding to be bundled with the exynos-simple-phy.h header.

> @@ -20,3 +20,4 @@ phy-exynos-usb2-$(CONFIG_PHY_EXYNOS4X12_USB2)	+= phy-exynos4x12-usb2.o
>  phy-exynos-usb2-$(CONFIG_PHY_EXYNOS5250_USB2)	+= phy-exynos5250-usb2.o
>  obj-$(CONFIG_PHY_EXYNOS5_USBDRD)	+= phy-exynos5-usbdrd.o
>  obj-$(CONFIG_PHY_XGENE)			+= phy-xgene.o
> +obj-$(CONFIG_EXYNOS_SIMPLE_PHY)		+= exynos-simple-phy.o

Perhaps this should be named phy-exynos-simple for consistency? Also it
may be a good idea to sort this alphabetically to reduce the potential
for conflicts.

> +static int exynos_phy_probe(struct platform_device *pdev)
> +{
> +	const struct of_device_id *of_id = of_match_device(
> +		of_match_ptr(exynos_phy_of_match), &pdev->dev);

Why does this need of_match_ptr()?

> +	dev_info(dev, "probe success\n");

If at all this should be dev_dbg(). But in general the driver core will
already complain if the driver fails to probe, so there's in general no
need to mention when it probes successfully.

> diff --git a/include/dt-bindings/phy/exynos-simple-phy.h b/include/dt-bindings/phy/exynos-simple-phy.h
[...]
> +/* simeple phys */

s/simeple phys/simple PHYs/

Although on second thought that comment probably shouldn't be there in
the first place.

> +#define INVALID	(~1)

This doesn't belong in this header. The value should never be used by a
DT source file, should it?

> +#define HDMI_PHY	0
> +#define DAC_PHY	1
> +#define ADC_PHY	2
> +#define PCIE_PHY	3
> +#define SATA_PHY	4

Perhaps these should be namespaced somehow to avoid potential conflicts
with other PHY providers?

> +#define PHY_NR	5

I'm not sure that this belongs here either. It's not a value that will
ever appear in a DT source file.

Thierry

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]