[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Yoshihiro Kaneko]

From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>

HSUSB uses hp_clk for H/W register access.
HSUSB uses mp_clk for USB-DMAC access.
The clock is defined for H/W register access.

Signed-off-by: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>
---

This patch is based on devel branch of Simon Horman's renesas tree.

 arch/arm/boot/dts/r8a7790.dtsi | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/r8a7790.dtsi b/arch/arm/boot/dts/r8a7790.dtsi
index 24de994..113edf2 100644
--- a/arch/arm/boot/dts/r8a7790.dtsi
+++ b/arch/arm/boot/dts/r8a7790.dtsi
@@ -1,6 +1,7 @@
 /*
  * Device Tree Source for the r8a7790 SoC
  *
+ * Copyright (C) 2015 Renesas Electronics Corporation
  * Copyright (C) 2013-2014 Renesas Solutions Corp.
  * Copyright (C) 2014 Cogent Embedded Inc.
  *
@@ -1132,7 +1133,7 @@
 		mstp7_clks: mstp7_clks at e615014c {
 			compatible = "renesas,r8a7790-mstp-clocks", "renesas,cpg-mstp-clocks";
 			reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
-			clocks = <&mp_clk>, <&mp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
+			clocks = <&mp_clk>, <&hp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
 				 <&p_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>,
 				 <&zx_clk>;
 			#clock-cells = <1>;
-- 
1.9.1

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Simon Horman]

On Sun, Feb 15, 2015 at 11:49:05PM +0900, Yoshihiro Kaneko wrote:
> From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> 
> HSUSB uses hp_clk for H/W register access.
> HSUSB uses mp_clk for USB-DMAC access.
> The clock is defined for H/W register access.
> 
> Signed-off-by: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>

I plan to take this through the renesas tree unless someone objects.
In the mean time, for reference:

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

> ---
> 
> This patch is based on devel branch of Simon Horman's renesas tree.
> 
>  arch/arm/boot/dts/r8a7790.dtsi | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/boot/dts/r8a7790.dtsi b/arch/arm/boot/dts/r8a7790.dtsi
> index 24de994..113edf2 100644
> --- a/arch/arm/boot/dts/r8a7790.dtsi
> +++ b/arch/arm/boot/dts/r8a7790.dtsi
> @@ -1,6 +1,7 @@
>  /*
>   * Device Tree Source for the r8a7790 SoC
>   *
> + * Copyright (C) 2015 Renesas Electronics Corporation
>   * Copyright (C) 2013-2014 Renesas Solutions Corp.
>   * Copyright (C) 2014 Cogent Embedded Inc.
>   *
> @@ -1132,7 +1133,7 @@
>  		mstp7_clks: mstp7_clks at e615014c {
>  			compatible = "renesas,r8a7790-mstp-clocks", "renesas,cpg-mstp-clocks";
>  			reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
> -			clocks = <&mp_clk>, <&mp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
> +			clocks = <&mp_clk>, <&hp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
>  				 <&p_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>,
>  				 <&zx_clk>;
>  			#clock-cells = <1>;
> -- 
> 1.9.1
> 

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Geert Uytterhoeven]

Hi Kaneko-san, Mizuguchi-san,

On Sun, Feb 15, 2015 at 3:49 PM, Yoshihiro Kaneko <ykaneko0929@gmail.com> wrote:
> From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
>
> HSUSB uses hp_clk for H/W register access.

That indeed seems to be true, and that's what this patch is changing.

> HSUSB uses mp_clk for USB-DMAC access.

However, page 64-2 doesn't mention the MP clock for USB-DMAC:
   "AXI bus clock (ZS?: 260 MHz), HPB bus clock (HP?: 130 MHz).
    USB local bus clock (48 MHz)"
So your statement doesn't seem to be true?

Note that mstp3_clks in dtsi has hp_clk as the parent for both usbdmac0 and
usbdmac1, so if your statement would be true, that has to changed, too?

The same is true for other members of the R-Car Gen2 family, e.g. your patch
for r8a7791.

> The clock is defined for H/W register access.
>
> Signed-off-by: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>
> ---
>
> This patch is based on devel branch of Simon Horman's renesas tree.
>
>  arch/arm/boot/dts/r8a7790.dtsi | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/arch/arm/boot/dts/r8a7790.dtsi b/arch/arm/boot/dts/r8a7790.dtsi
> index 24de994..113edf2 100644
> --- a/arch/arm/boot/dts/r8a7790.dtsi
> +++ b/arch/arm/boot/dts/r8a7790.dtsi

> @@ -1132,7 +1133,7 @@
>                 mstp7_clks: mstp7_clks at e615014c {
>                         compatible = "renesas,r8a7790-mstp-clocks", "renesas,cpg-mstp-clocks";
>                         reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
> -                       clocks = <&mp_clk>, <&mp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
> +                       clocks = <&mp_clk>, <&hp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
>                                  <&p_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>,
>                                  <&zx_clk>;
>                         #clock-cells = <1>;

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at 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

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Simon Horman]

On Mon, Feb 16, 2015 at 10:20:10AM +0100, Geert Uytterhoeven wrote:
> Hi Kaneko-san, Mizuguchi-san,
> 
> On Sun, Feb 15, 2015 at 3:49 PM, Yoshihiro Kaneko <ykaneko0929@gmail.com> wrote:
> > From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> >
> > HSUSB uses hp_clk for H/W register access.
> 
> That indeed seems to be true, and that's what this patch is changing.
> 
> > HSUSB uses mp_clk for USB-DMAC access.
> 
> However, page 64-2 doesn't mention the MP clock for USB-DMAC:
>    "AXI bus clock (ZS?: 260 MHz), HPB bus clock (HP?: 130 MHz).
>     USB local bus clock (48 MHz)"
> So your statement doesn't seem to be true?

Thanks, I see that too.

> Note that mstp3_clks in dtsi has hp_clk as the parent for both usbdmac0 and
> usbdmac1, so if your statement would be true, that has to changed, too?
> 
> The same is true for other members of the R-Car Gen2 family, e.g. your patch
> for r8a7791.

It seems to me that we could:

1. Apply this patch with the changelog updated and;
2. Follow up on the DMAC clocks

Does that sound reasonable to you?

> > The clock is defined for H/W register access.
> >
> > Signed-off-by: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> > Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>
> > ---
> >
> > This patch is based on devel branch of Simon Horman's renesas tree.
> >
> >  arch/arm/boot/dts/r8a7790.dtsi | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/arch/arm/boot/dts/r8a7790.dtsi b/arch/arm/boot/dts/r8a7790.dtsi
> > index 24de994..113edf2 100644
> > --- a/arch/arm/boot/dts/r8a7790.dtsi
> > +++ b/arch/arm/boot/dts/r8a7790.dtsi
> 
> > @@ -1132,7 +1133,7 @@
> >                 mstp7_clks: mstp7_clks at e615014c {
> >                         compatible = "renesas,r8a7790-mstp-clocks", "renesas,cpg-mstp-clocks";
> >                         reg = <0 0xe615014c 0 4>, <0 0xe61501c4 0 4>;
> > -                       clocks = <&mp_clk>, <&mp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
> > +                       clocks = <&mp_clk>, <&hp_clk>, <&zs_clk>, <&zs_clk>, <&p_clk>,
> >                                  <&p_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>, <&zx_clk>,
> >                                  <&zx_clk>;
> >                         #clock-cells = <1>;
> 
> Gr{oetje,eeting}s,
> 
>                         Geert
> 
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at 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
> 

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Geert Uytterhoeven]

Hi Simon,

On Mon, Feb 16, 2015 at 2:57 PM, Simon Horman <horms@verge.net.au> wrote:
> On Mon, Feb 16, 2015 at 10:20:10AM +0100, Geert Uytterhoeven wrote:
>> Hi Kaneko-san, Mizuguchi-san,
>>
>> On Sun, Feb 15, 2015 at 3:49 PM, Yoshihiro Kaneko <ykaneko0929@gmail.com> wrote:
>> > From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
>> >
>> > HSUSB uses hp_clk for H/W register access.
>>
>> That indeed seems to be true, and that's what this patch is changing.
>>
>> > HSUSB uses mp_clk for USB-DMAC access.
>>
>> However, page 64-2 doesn't mention the MP clock for USB-DMAC:
>>    "AXI bus clock (ZS?: 260 MHz), HPB bus clock (HP?: 130 MHz).
>>     USB local bus clock (48 MHz)"
>> So your statement doesn't seem to be true?
>
> Thanks, I see that too.
>
>> Note that mstp3_clks in dtsi has hp_clk as the parent for both usbdmac0 and
>> usbdmac1, so if your statement would be true, that has to changed, too?
>>
>> The same is true for other members of the R-Car Gen2 family, e.g. your patch
>> for r8a7791.
>
> It seems to me that we could:
>
> 1. Apply this patch with the changelog updated and;
> 2. Follow up on the DMAC clocks
>
> Does that sound reasonable to you?

Yes, that sounds fine to me.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at 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

[PATCH/RFC] ARM: shmobile: r8a7790: Fix HSUSB clock to hp_clk from mp_clk

[Simon Horman]

Hi Geert,

On Mon, Feb 16, 2015 at 03:04:16PM +0100, Geert Uytterhoeven wrote:
> Hi Simon,
> 
> On Mon, Feb 16, 2015 at 2:57 PM, Simon Horman <horms@verge.net.au> wrote:
> > On Mon, Feb 16, 2015 at 10:20:10AM +0100, Geert Uytterhoeven wrote:
> >> Hi Kaneko-san, Mizuguchi-san,
> >>
> >> On Sun, Feb 15, 2015 at 3:49 PM, Yoshihiro Kaneko <ykaneko0929@gmail.com> wrote:
> >> > From: Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
> >> >
> >> > HSUSB uses hp_clk for H/W register access.
> >>
> >> That indeed seems to be true, and that's what this patch is changing.
> >>
> >> > HSUSB uses mp_clk for USB-DMAC access.
> >>
> >> However, page 64-2 doesn't mention the MP clock for USB-DMAC:
> >>    "AXI bus clock (ZS?: 260 MHz), HPB bus clock (HP?: 130 MHz).
> >>     USB local bus clock (48 MHz)"
> >> So your statement doesn't seem to be true?
> >
> > Thanks, I see that too.
> >
> >> Note that mstp3_clks in dtsi has hp_clk as the parent for both usbdmac0 and
> >> usbdmac1, so if your statement would be true, that has to changed, too?
> >>
> >> The same is true for other members of the R-Car Gen2 family, e.g. your patch
> >> for r8a7791.
> >
> > It seems to me that we could:
> >
> > 1. Apply this patch with the changelog updated and;
> > 2. Follow up on the DMAC clocks
> >
> > Does that sound reasonable to you?
> 
> Yes, that sounds fine to me.

The purpose of this email is to offer a quick status update which
I hope will bring this matter to a close.

I have submitted and queued up a patch as per 1. above.

With regards to item 2. above: I have also followed up on the "HSUSB uses
mp_clk for USB-DMAC access" comment in the original changelog (above).
Mizuguchi-san has confirmed that statement is not correct and its presence
in the changelog was an error.