[PATCH 1/2] ARM: dts: armada388-clearfog: increase speed of i2c0 to 400kHz

[Gregory CLEMENT]

Hi Russell King,
 
 On lun., mars 05 2018, Russell King - ARM Linux <linux at armlinux.org.uk> wrote:

> On Mon, Mar 05, 2018 at 04:48:54PM +0100, Gregory CLEMENT wrote:
>> Hi Russell,
>>  
>>  On mar., févr. 27 2018, Russell King <rmk+kernel at armlinux.org.uk> wrote:
>> 
>> > All the devices on I2C0 support fast mode, so increase the bus speed
>> > to match.
>> 
>> I agree with this part. I didn't pay attention to it in the initial
>> submission but now I've checked it, and indeed according to the
>> datasheet the MCP3021 support "standard (100 kHz) and Fast (400 kHz) I2C
>> Compatible modes"
>> 
>> > The Armada 388 is known to have a timing issue when in
>> > standard mode, which we believe causes the ficticious device at 0x64
>> > to appear.
>> 
>> This errata "FE-8471889 I2C Repeated Start is not supported in standard
>> mode (up to 100K)" was supposed to be fixed with the commit c1d15b68aab8
>> ("i2c: mv64xxx: Fix timing issue on Armada XP (errata FE-8471889)"). And
>> it was used on the Armada 38x SoCs since fbffee74986c ("ARM: dts: Fix I2C
>> repeated start issue on Armada-38x") in 4.14.
>
> The commit pre-dates 4.14, and the commit message was correct at the time
> it was committed (Nov 29 2016).  I don't know whether its still the case
> that the ficticious device still appears, but increasing the bus speed is
> still a sensible thing to do if the devices all support it.

I agree to apply it. I was concerned by the fact that the i2c fix might
not work. I've just tested it and by using the "marvell,mv78230-a0-i2c"
there was no more ficticious device at 100KHz, so I am happy with it.

In order to make know that now the Standard mode is working on Armada
38x, I will just add a the following comment to your commit log:

[gregory.clement at bootlin.com: Note that since the commit fbffee74986c
("ARM: dts: Fix I2C repeated start issue on Armada-38x") in 4.14, the
timing issue is managed for the Armada 38x SoCs.]

Thanks,

Gregory

>
>> So I think that now even at 100kHz the ficticious device should not
>> appear any more. If it is still the case then there might be an issue
>> with the fix in the 2c driver.
>> 
>> Gregory
>> 
>> >
>> > Signed-off-by: Russell King <rmk+kernel at armlinux.org.uk>
>> > ---
>> >  arch/arm/boot/dts/armada-388-clearfog.dtsi | 7 ++-----
>> >  1 file changed, 2 insertions(+), 5 deletions(-)
>> >
>> > diff --git a/arch/arm/boot/dts/armada-388-clearfog.dtsi b/arch/arm/boot/dts/armada-388-clearfog.dtsi
>> > index 68acfc968706..0e3b1f140e6d 100644
>> > --- a/arch/arm/boot/dts/armada-388-clearfog.dtsi
>> > +++ b/arch/arm/boot/dts/armada-388-clearfog.dtsi
>> > @@ -143,8 +143,7 @@
>> >  };
>> >  
>> >  &i2c0 {
>> > -	/* Is there anything on this? */
>> > -	clock-frequency = <100000>;
>> > +	clock-frequency = <400000>;
>> >  	pinctrl-0 = <&i2c0_pins>;
>> >  	pinctrl-names = "default";
>> >  	status = "okay";
>> > @@ -239,13 +238,11 @@
>> >  		};
>> >  	};
>> >  
>> > -	/* The MCP3021 is 100kHz clock only */
>> > +	/* The MCP3021 supports standard and fast modes */
>> >  	mikrobus_adc: mcp3021 at 4c {
>> >  		compatible = "microchip,mcp3021";
>> >  		reg = <0x4c>;
>> >  	};
>> > -
>> > -	/* Also something at 0x64 */
>> >  };
>> >  
>> >  &i2c1 {
>> > -- 
>> > 2.7.4
>> >
>> 
>> -- 
>> Gregory Clement, Bootlin (formerly Free Electrons)
>> Embedded Linux and Kernel engineering
>> http://bootlin.com
>
> -- 
> RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
> FTTC broadband for 0.8mile line in suburbia: sync at 8.8Mbps down 630kbps up
> According to speedtest.net: 8.21Mbps down 510kbps up

-- 
Gregory Clement, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com