From mboxrd@z Thu Jan  1 00:00:00 1970
From: sebastian.hesselbarth@gmail.com (Sebastian Hesselbarth)
Date: Fri, 21 Mar 2014 13:11:29 +0100
Subject: [PATCH 3/5] ARM: berlin/dt: add cpupll and syspll support to BG2Q
In-Reply-To: <1395402220-23503-4-git-send-email-alexandre.belloni@free-electrons.com>
References: <1395402220-23503-1-git-send-email-alexandre.belloni@free-electrons.com>
 <1395402220-23503-4-git-send-email-alexandre.belloni@free-electrons.com>
Message-ID: <532C2C71.4010703@gmail.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On 03/21/2014 12:43 PM, Alexandre Belloni wrote:

Alexandre,

Thanks for starting this! I'll start with the most obvious
things first and have a closer look on it later.

Missing commit description here.

> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
> ---
>   arch/arm/boot/dts/berlin2q.dtsi | 20 ++++++++++++++++++--
>   1 file changed, 18 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm/boot/dts/berlin2q.dtsi b/arch/arm/boot/dts/berlin2q.dtsi
> index 07452a7483fa..19d2c82b0664 100644
> --- a/arch/arm/boot/dts/berlin2q.dtsi
> +++ b/arch/arm/boot/dts/berlin2q.dtsi
> @@ -59,10 +59,26 @@
>   		clock-frequency = <100000000>;
>   	};
>
> +	syspll: syspll {

syspll: pll at ea0030 {

and sort it in between other SoC nodes below. This will
most likely break clocks in v3.14 but v3.15 will receive
proper clock init ordering.

> +		compatible = "marvell,berlin2q-pll";
> +		clocks = <&smclk>;
> +		#clock-cells = <0>;
> +		reg = <0xf7ea0030 8>;
> +	};
> +
> +	cpupll: cpupll {

dito.

> +		compatible = "marvell,berlin2q-pll";
> +		clocks = <&smclk>;
> +		#clock-cells = <0>;
> +		reg = <0xf7dd0170 8>;
> +	};
> +
>   	cpuclk: cpu-clock {
> -		compatible = "fixed-clock";
> +		compatible = "fixed-factor-clock";
> +		clocks = <&cpupll>;
>   		#clock-cells = <0>;
> -		clock-frequency = <1200000000>;
> +		clock-div = <1>;
> +		clock-mult = <1>;

Hmm, you probably know better than me, but if cpuclk == cpupll
is always true we don't need another clk layer here. If you
can scale down cpuclk from cpupll and we just have no driver
for it, I am fine with it.

Sebastian

>   	};
>
>   	twdclk: twdclk {
>