[PATCH/RFC] ARM: shmobile: r8a7794: Add Audio DMAC, PWM and Thermal clocks to device tree

[PATCH/RFC] ARM: shmobile: r8a7794: Add Audio DMAC, PWM and Thermal clocks to device tree

[Simon Horman]

This is based on the MSTP5 clocks in the r8a7791 device tree.
The main difference being that the r8a7794 does not have an Audio DMAC1
clock.

Signed-off-by: Simon Horman <horms+renesas@verge.net.au>

---
Based on the renesas-devel-20141217-v3.18 tag of my renesas tree.

N.B: The R-Car Gen2 v1.8.0 BSP uses zs_clk whereas this patch uses hp_clk
     as the parent clock for the Audio DMAC clock.

     In that regard this patch follows the pattern used in mainline for the
     r8a7790 and r8a7791 SoCs.

     The BSP's parent clock is consistent for the r8a7790, r8a7791, r8a7793
     and r8a7794 SoCs.

     Support for the r8a7793 SoC is not present in mainline.
---
 arch/arm/boot/dts/r8a7794.dtsi            | 11 +++++++++++
 include/dt-bindings/clock/r8a7794-clock.h |  1 +
 2 files changed, 12 insertions(+)

diff --git a/arch/arm/boot/dts/r8a7794.dtsi b/arch/arm/boot/dts/r8a7794.dtsi
index 063bf56..7a03166 100644
--- a/arch/arm/boot/dts/r8a7794.dtsi
+++ b/arch/arm/boot/dts/r8a7794.dtsi
@@ -530,6 +530,17 @@
 			        "sdhi2", "sdhi1", "sdhi0",
 				"mmcif0", "cmt1", "usbdmac0", "usbdmac1";
 		};
+		mstp5_clks: mstp5_clks@e6150144 {
+			compatible = "renesas,r8a7794-mstp-clocks", "renesas,cpg-mstp-clocks";
+			reg = <0 0xe6150144 0 4>, <0 0xe615003c 0 4>;
+			clocks = <&hp_clk>, <&extal_clk>, <&p_clk>;
+			#clock-cells = <1>;
+			clock-indices = <
+				R8A7794_CLK_AUDIO_DMAC0
+				R8A7794_CLK_THERMAL R8A7794_CLK_PWM
+			>;
+			clock-output-names = "audmac0", "thermal", "pwm";
+		};
 		mstp7_clks: mstp7_clks@e615014c {
 			compatible = "renesas,r8a7794-mstp-clocks", "renesas,cpg-mstp-clocks";
 			reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
diff --git a/include/dt-bindings/clock/r8a7794-clock.h b/include/dt-bindings/clock/r8a7794-clock.h
index d5cfe62..58fa44d 100644
--- a/include/dt-bindings/clock/r8a7794-clock.h
+++ b/include/dt-bindings/clock/r8a7794-clock.h
@@ -61,6 +61,7 @@
 #define R8A7794_CLK_USBDMAC1		31
 
 /* MSTP5 */
+#define R8A7794_CLK_AUDIO_DMAC0		2
 #define R8A7794_CLK_THERMAL		22
 #define R8A7794_CLK_PWM			23
 
-- 
2.1.3

Re: [PATCH/RFC] ARM: shmobile: r8a7794: Add Audio DMAC, PWM and Thermal clocks to device tree

[Laurent Pinchart]

Hi Simon,

On Friday 19 December 2014 09:39:33 Simon Horman wrote:
> This is based on the MSTP5 clocks in the r8a7791 device tree.
> The main difference being that the r8a7794 does not have an Audio DMAC1
> clock.
> 
> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
> 
> ---
> Based on the renesas-devel-20141217-v3.18 tag of my renesas tree.
> 
> N.B: The R-Car Gen2 v1.8.0 BSP uses zs_clk whereas this patch uses hp_clk
>      as the parent clock for the Audio DMAC clock.
> 
>      In that regard this patch follows the pattern used in mainline for the
>      r8a7790 and r8a7791 SoCs.
> 
>      The BSP's parent clock is consistent for the r8a7790, r8a7791, r8a7793
>      and r8a7794 SoCs.

Does this mean the BSP uses hp_clk for all Gen2 SoCs while mainline uses 
zs_clk ? Can't we find out which one is correct ?

>      Support for the r8a7793 SoC is not present in mainline.
> ---
>  arch/arm/boot/dts/r8a7794.dtsi            | 11 +++++++++++
>  include/dt-bindings/clock/r8a7794-clock.h |  1 +
>  2 files changed, 12 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/r8a7794.dtsi b/arch/arm/boot/dts/r8a7794.dtsi
> index 063bf56..7a03166 100644
> --- a/arch/arm/boot/dts/r8a7794.dtsi
> +++ b/arch/arm/boot/dts/r8a7794.dtsi
> @@ -530,6 +530,17 @@
>  			        "sdhi2", "sdhi1", "sdhi0",
>  				"mmcif0", "cmt1", "usbdmac0", "usbdmac1";
>  		};
> +		mstp5_clks: mstp5_clks@e6150144 {
> +			compatible = "renesas,r8a7794-mstp-clocks", "renesas,cpg-mstp-
clocks";
> +			reg = <0 0xe6150144 0 4>, <0 0xe615003c 0 4>;
> +			clocks = <&hp_clk>, <&extal_clk>, <&p_clk>;
> +			#clock-cells = <1>;
> +			clock-indices = <
> +				R8A7794_CLK_AUDIO_DMAC0
> +				R8A7794_CLK_THERMAL R8A7794_CLK_PWM
> +			>;
> +			clock-output-names = "audmac0", "thermal", "pwm";
> +		};
>  		mstp7_clks: mstp7_clks@e615014c {
>  			compatible = "renesas,r8a7794-mstp-clocks", "renesas,cpg-mstp-
clocks";
>  			reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
> diff --git a/include/dt-bindings/clock/r8a7794-clock.h
> b/include/dt-bindings/clock/r8a7794-clock.h index d5cfe62..58fa44d 100644
> --- a/include/dt-bindings/clock/r8a7794-clock.h
> +++ b/include/dt-bindings/clock/r8a7794-clock.h
> @@ -61,6 +61,7 @@
>  #define R8A7794_CLK_USBDMAC1		31
> 
>  /* MSTP5 */
> +#define R8A7794_CLK_AUDIO_DMAC0		2
>  #define R8A7794_CLK_THERMAL		22
>  #define R8A7794_CLK_PWM			23

-- 
Regards,

Laurent Pinchart

Re: [PATCH/RFC] ARM: shmobile: r8a7794: Add Audio DMAC, PWM and Thermal clocks to device tree

[Geert Uytterhoeven]

On Mon, Jan 5, 2015 at 9:30 AM, Laurent Pinchart
<laurent.pinchart@ideasonboard.com> wrote:
> On Friday 19 December 2014 09:39:33 Simon Horman wrote:
>> This is based on the MSTP5 clocks in the r8a7791 device tree.
>> The main difference being that the r8a7794 does not have an Audio DMAC1
>> clock.
>>
>> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
>>
>> ---
>> Based on the renesas-devel-20141217-v3.18 tag of my renesas tree.
>>
>> N.B: The R-Car Gen2 v1.8.0 BSP uses zs_clk whereas this patch uses hp_clk
>>      as the parent clock for the Audio DMAC clock.
>>
>>      In that regard this patch follows the pattern used in mainline for the
>>      r8a7790 and r8a7791 SoCs.
>>
>>      The BSP's parent clock is consistent for the r8a7790, r8a7791, r8a7793
>>      and r8a7794 SoCs.
>
> Does this mean the BSP uses hp_clk for all Gen2 SoCs while mainline uses
> zs_clk ?

No, it's the other way around ;-)

> Can't we find out which one is correct ?

Which clock is used doesn't seem to be documented.
So either we have to check with the hardware guys, or measure the clock
in some way, either by hardware or software (e.g. what's the maximum
transfer rate
that works?).

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH/RFC] ARM: shmobile: r8a7794: Add Audio DMAC, PWM and Thermal clocks to device tree

[Laurent Pinchart]

Hi Geert,

On Monday 05 January 2015 09:52:37 Geert Uytterhoeven wrote:
> On Mon, Jan 5, 2015 at 9:30 AM, Laurent Pinchart wrote:
> > On Friday 19 December 2014 09:39:33 Simon Horman wrote:
> >> This is based on the MSTP5 clocks in the r8a7791 device tree.
> >> The main difference being that the r8a7794 does not have an Audio DMAC1
> >> clock.
> >> 
> >> Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
> >> 
> >> ---
> >> Based on the renesas-devel-20141217-v3.18 tag of my renesas tree.
> >> 
> >> N.B: The R-Car Gen2 v1.8.0 BSP uses zs_clk whereas this patch uses hp_clk
> >>      as the parent clock for the Audio DMAC clock.
> >>      
> >>      In that regard this patch follows the pattern used in mainline for
> >>      the r8a7790 and r8a7791 SoCs.
> >>      
> >>      The BSP's parent clock is consistent for the r8a7790, r8a7791,
> >>      r8a7793 and r8a7794 SoCs.
> > 
> > Does this mean the BSP uses hp_clk for all Gen2 SoCs while mainline uses
> > zs_clk ?
> 
> No, it's the other way around ;-)

Sorry, that's what I meant.

> > Can't we find out which one is correct ?
> 
> Which clock is used doesn't seem to be documented.
> So either we have to check with the hardware guys, or measure the clock
> in some way, either by hardware or software (e.g. what's the maximum
> transfer rate that works?).

Given the parent clock frequencies (130 MHz and 260 MHz) and the audio rates, 
I'm not sure any useful measurement could be performed.

-- 
Regards,

Laurent Pinchart