[PATCH] arm: dts: gr-peach: Reduce extal_clk resolution

[PATCH] arm: dts: gr-peach: Reduce extal_clk resolution

[Jacopo Mondi]

The system clock described by extal_clk is reported to have a frequency
of 13.333 Mhz and is correctly described by gr-peach device tree.

However, when enabling a RIIC device the following error is reported by
drivers/i2c/busses/i2c-riic.c

"invalid parent clk (33332500). Must be 33325000Hz"

As RIIC devices have a clock source obtained by dividing by 12 the
system clock, the resulting value is not accepted by the driver
(which clearly states not to support any frequency except the reported
33325000Hz one).

Hence, reduce the system clock accuracy to a value which makes
frequencies obtained through division accepted by RIIC driver.

Please note that other r7s72100 boards, such as Genmai, report the same
"reduced accuracy" frequency, even if their external clock sources are
effectively 13.333Mhz as gr-peach one.

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
---
 arch/arm/boot/dts/r7s72100-gr-peach.dts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/r7s72100-gr-peach.dts b/arch/arm/boot/dts/r7s72100-gr-peach.dts
index a1c5e88..20309ac 100644
--- a/arch/arm/boot/dts/r7s72100-gr-peach.dts
+++ b/arch/arm/boot/dts/r7s72100-gr-peach.dts
@@ -71,7 +71,7 @@
 };

 &extal_clk {
-	clock-frequency = <13333000>;
+	clock-frequency = <13330000>;
 };

 &usb_x1_clk {
--
2.7.4

[PATCH] arm: dts: gr-peach: Reduce extal_clk resolution

[Geert Uytterhoeven]

Hi Jacopo,

On Wed, Oct 4, 2017 at 5:40 PM, Jacopo Mondi <jacopo+renesas@jmondi.org> wrote:
> The system clock described by extal_clk is reported to have a frequency
> of 13.333 Mhz and is correctly described by gr-peach device tree.
>
> However, when enabling a RIIC device the following error is reported by
> drivers/i2c/busses/i2c-riic.c
>
> "invalid parent clk (33332500). Must be 33325000Hz"
>
> As RIIC devices have a clock source obtained by dividing by 12 the
> system clock, the resulting value is not accepted by the driver
> (which clearly states not to support any frequency except the reported
> 33325000Hz one).
>
> Hence, reduce the system clock accuracy to a value which makes
> frequencies obtained through division accepted by RIIC driver.
>
> Please note that other r7s72100 boards, such as Genmai, report the same
> "reduced accuracy" frequency, even if their external clock sources are
> effectively 13.333Mhz as gr-peach one.
>
> Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>

There's no need to do this, Chris already proposed a fix, cfr.
"[PATCH v2] i2c: riic: remove fixed clock restriction"
(https://www.spinics.net/lists/linux-renesas-soc/msg18573.html).

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] arm: dts: gr-peach: Reduce extal_clk resolution

[jacopo mondi]

Hi Geert

On Wed, Oct 04, 2017 at 05:54:46PM +0200, Geert Uytterhoeven wrote:
> Hi Jacopo,
>
> On Wed, Oct 4, 2017 at 5:40 PM, Jacopo Mondi <jacopo+renesas@jmondi.org> wrote:
> > The system clock described by extal_clk is reported to have a frequency
> > of 13.333 Mhz and is correctly described by gr-peach device tree.
> >
> > However, when enabling a RIIC device the following error is reported by
> > drivers/i2c/busses/i2c-riic.c
> >
> > "invalid parent clk (33332500). Must be 33325000Hz"
> >
> > As RIIC devices have a clock source obtained by dividing by 12 the
> > system clock, the resulting value is not accepted by the driver
> > (which clearly states not to support any frequency except the reported
> > 33325000Hz one).
> >
> > Hence, reduce the system clock accuracy to a value which makes
> > frequencies obtained through division accepted by RIIC driver.
> >
> > Please note that other r7s72100 boards, such as Genmai, report the same
> > "reduced accuracy" frequency, even if their external clock sources are
> > effectively 13.333Mhz as gr-peach one.
> >
> > Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
>
> There's no need to do this, Chris already proposed a fix, cfr.
> "[PATCH v2] i2c: riic: remove fixed clock restriction"
> (https://www.spinics.net/lists/linux-renesas-soc/msg18573.html).

Oh that's even better!
Do you see any value in augmenting the frequency resolution in Genmai
DTS then?

Thanks
  j

>
> 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] arm: dts: gr-peach: Reduce extal_clk resolution

[Geert Uytterhoeven]

Hi Jacopo,

On Thu, Oct 5, 2017 at 9:48 AM, jacopo mondi <jacopo@jmondi.org> wrote:
> On Wed, Oct 04, 2017 at 05:54:46PM +0200, Geert Uytterhoeven wrote:
>> On Wed, Oct 4, 2017 at 5:40 PM, Jacopo Mondi <jacopo+renesas@jmondi.org> wrote:
>> > The system clock described by extal_clk is reported to have a frequency
>> > of 13.333 Mhz and is correctly described by gr-peach device tree.
>> >
>> > However, when enabling a RIIC device the following error is reported by
>> > drivers/i2c/busses/i2c-riic.c
>> >
>> > "invalid parent clk (33332500). Must be 33325000Hz"
>> >
>> > As RIIC devices have a clock source obtained by dividing by 12 the
>> > system clock, the resulting value is not accepted by the driver
>> > (which clearly states not to support any frequency except the reported
>> > 33325000Hz one).
>> >
>> > Hence, reduce the system clock accuracy to a value which makes
>> > frequencies obtained through division accepted by RIIC driver.
>> >
>> > Please note that other r7s72100 boards, such as Genmai, report the same
>> > "reduced accuracy" frequency, even if their external clock sources are
>> > effectively 13.333Mhz as gr-peach one.
>> >
>> > Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
>>
>> There's no need to do this, Chris already proposed a fix, cfr.
>> "[PATCH v2] i2c: riic: remove fixed clock restriction"
>> (https://www.spinics.net/lists/linux-renesas-soc/msg18573.html).
>
> Oh that's even better!
> Do you see any value in augmenting the frequency resolution in Genmai
> DTS then?

The Genmai documentation says EXTAL is 13.33 MHz (which matches
r7s72100-genmai.dts) everywhere, except in the schematics, where it says
13.3333 MHz. Sigh...

The difference is slightly larger than 100 ppm, so still insignificant, I think.

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