Re: [PATCH V4 3/3] ARM: dts: Add pcie controller node for Samsung EXYNOS5440 SoC

[Arnd Bergmann <arnd-r2nGTMty4D4@public.gmane.org>]

Thanks for the update! A few comments again:

On Wednesday 12 June 2013 19:20:00 Jingoo Han wrote:
> 
> diff --git a/arch/arm/boot/dts/exynos5440-ssdk5440.dts b/arch/arm/boot/dts/exynos5440-ssdk5440.dts
> index d55042b..efe7d39 100644
> --- a/arch/arm/boot/dts/exynos5440-ssdk5440.dts
> +++ b/arch/arm/boot/dts/exynos5440-ssdk5440.dts
> @@ -30,4 +30,12 @@
>                         clock-frequency = <50000000>;
>                 };
>         };
> +
> +       pcie0@40000000 {
> +               reset-gpio = <5>;
> +       };
> +
> +       pcie1@60000000 {
> +               reset-gpio = <22>;
> +       };
>  };

As mentioned before, please use the gpio binding to pass gpio numbers.

> diff --git a/arch/arm/boot/dts/exynos5440.dtsi b/arch/arm/boot/dts/exynos5440.dtsi
> index f6b1c89..2c15f9d 100644
> --- a/arch/arm/boot/dts/exynos5440.dtsi
> +++ b/arch/arm/boot/dts/exynos5440.dtsi
> @@ -216,4 +216,42 @@
>                 clock-names = "rtc";
>                 status = "disabled";
>         };
> +
> +       pcie0@0x290000 {
> +               compatible = "samsung,exynos5440-pcie";
> +               reg = <0x290000 0x1000
> +                       0x270000 0x1000
> +                       0x271000 0x40>;
> +               interrupts = <0 20 0>, <0 21 0>, <0 22 0>;
> +               clocks = <&clock 28>, <&clock 27>;
> +               clock-names = "pcie", "pcie_bus";
> +               #address-cells = >;
> +               #size-cells = <2>;
> +               device_type = "pci";
> +               ranges = <0x00000800 0 0x40000000 0x40000000 0 0x00200000   /* configuration space */
> +                         0x81000000 0 0          0x40200000 0 0x00010000   /* downstream I/O */
> +                         0x82000000 0 0x40210000 0x40210000 0 0x10000000>; /* non-prefetchable memory */

I think you did not reply to my question regarding the size of the
memory space. Does it extend from 0x40210000 to 0x50210000,
or from 0x40210000 to 0x50000000. You probably meant the latter
but wrote the former. If not, please add a comment for clarification.

> +               #interrupt-cells = <1>;
> +               interrupt-map-mask = <0 0 0 0>;
> +               interrupt-map = <0x0 0 &gic 53>;
> +       };

So all PCI IntA interrupts are mapped to a single gic interrupt? That
sounds like a bottleneck when you have a lot of devices on the bus.
Do you have MSI support?

	Arnd