[PATCH 1/2] ARM: dts: imx8m: increase off-on delay on the SD Vcc regulator

[ZHIZHIKIN Andrey]

Hello Peter,

> -----Original Message-----
> From: Peter Robinson <pbrobinson at gmail.com>
> Sent: Wednesday, December 2, 2020 9:45 AM
> To: ZHIZHIKIN Andrey <andrey.zhizhikin at leica-geosystems.com>
> Cc: u-boot at lists.denx.de; sbabic at denx.de
> Subject: Re: [PATCH 1/2] ARM: dts: imx8m: increase off-on delay on the SD Vcc
> regulator
> 
> 
> On Tue, Dec 1, 2020 at 1:31 PM Andrey Zhizhikin <andrey.zhizhikin at leica-
> geosystems.com> wrote:
> >
> > Some SD Card controller and power circuitry has increased capacitance,
> > which keeps the internal logic remains powered after regulator is
> > switch off. This is generally the case when card is switched to SD104
> > mode, where a power cycle should be performed. In case if the card
> > internal logic remains powered, it causes a subsequent failure of mode
> > transition, effectively leading to failed enumeration.
> >
> > Introduce a delay of 20 msec in order to provide a possibility for
> > internal card circuitry to drain voltages and perform a power cycle
> > correctly.
> >
> > Similar fix is done in commit c49d0ac38a76 ("ARM: dts: rmobile:
> > Increase off-on delay on the SD Vcc regulator") targeted Renesas SOCs.
> >
> > Signed-off-by: Andrey Zhizhikin
> > <andrey.zhizhikin at leica-geosystems.com>
> > Cc: Stefano Babic <sbabic at denx.de>
> > ---
> >  arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi | 4 ++++
> >  arch/arm/dts/imx8mm-evk-u-boot.dtsi        | 4 ++++
> >  arch/arm/dts/imx8mn-ddr4-evk-u-boot.dtsi   | 4 ++++
> >  arch/arm/dts/imx8mp-evk-u-boot.dtsi        | 4 ++++
> >  arch/arm/dts/imx8mq-evk.dts                | 1 +
> >  arch/arm/dts/imx8mq-phanbell.dts           | 1 +
> >  6 files changed, 18 insertions(+)
> >
> > diff --git a/arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi
> > b/arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi
> > index fc1aebb2fe..6d80a529ae 100644
> > --- a/arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi
> > +++ b/arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi
> > @@ -37,6 +37,10 @@
> >         /delete-property/ assigned-clock-rates;  };
> >
> > +&reg_usdhc2_vmmc {
> > +       u-boot,off-on-delay-us = <20000>; };
> > +
> >  &fec1 {
> >         phy-reset-gpios = <&gpio4 22 GPIO_ACTIVE_LOW>;  }; diff --git
> > a/arch/arm/dts/imx8mm-evk-u-boot.dtsi
> > b/arch/arm/dts/imx8mm-evk-u-boot.dtsi
> > index b5c12105a9..9f77d3c6ff 100644
> > --- a/arch/arm/dts/imx8mm-evk-u-boot.dtsi
> > +++ b/arch/arm/dts/imx8mm-evk-u-boot.dtsi
> > @@ -46,6 +46,10 @@
> >         u-boot,dm-spl;
> >  };
> >
> > +&reg_usdhc2_vmmc {
> > +       u-boot,off-on-delay-us = <20000>; };
> > +
> >  &pinctrl_reg_usdhc2_vmmc {
> >         u-boot,dm-spl;
> >  };
> > diff --git a/arch/arm/dts/imx8mn-ddr4-evk-u-boot.dtsi
> > b/arch/arm/dts/imx8mn-ddr4-evk-u-boot.dtsi
> > index 4419679d4c..98b0b9891b 100644
> > --- a/arch/arm/dts/imx8mn-ddr4-evk-u-boot.dtsi
> > +++ b/arch/arm/dts/imx8mn-ddr4-evk-u-boot.dtsi
> > @@ -47,6 +47,10 @@
> >         u-boot,dm-spl;
> >  };
> >
> > +&reg_usdhc2_vmmc {
> > +       u-boot,off-on-delay-us = <20000>; };
> > +
> >  &pinctrl_uart2 {
> >         u-boot,dm-spl;
> >  };
> > diff --git a/arch/arm/dts/imx8mp-evk-u-boot.dtsi
> > b/arch/arm/dts/imx8mp-evk-u-boot.dtsi
> > index 24a93ac2d6..2452e9175c 100644
> > --- a/arch/arm/dts/imx8mp-evk-u-boot.dtsi
> > +++ b/arch/arm/dts/imx8mp-evk-u-boot.dtsi
> > @@ -48,6 +48,10 @@
> >         u-boot,dm-spl;
> >  };
> >
> > +&reg_usdhc2_vmmc {
> > +       u-boot,off-on-delay-us = <20000>; };
> > +
> >  &reg_usdhc2_vmmc {
> >         u-boot,dm-spl;
> >  };
> > diff --git a/arch/arm/dts/imx8mq-evk.dts b/arch/arm/dts/imx8mq-evk.dts
> > index 55294ba9c8..9663683f69 100644
> > --- a/arch/arm/dts/imx8mq-evk.dts
> > +++ b/arch/arm/dts/imx8mq-evk.dts
> > @@ -39,6 +39,7 @@
> >                 regulator-max-microvolt = <3300000>;
> >                 gpio = <&gpio2 19 GPIO_ACTIVE_HIGH>;
> >                 enable-active-high;
> > +               u-boot,off-on-delay-us = <20000>;
> 
> These should be going in the -u-boot.dtsi like the devices above. Why the two
> different ways of doing it?

imx8mq-evk-u-boot.dtsi is not provided in the source tree. Do you suggest I
should create one and move those bindings there?

> Also if all devices based on the imx8m[mnpq ] need
> this change it might be better putting it in the imx8m[mnpq]-u-boot.dtsi so any
> new boards that get added, like the recently posted imx8mm IoT Gate
> automatically benefit from this fix.

This might depend on the layout of the particular board. Some boards might include
additional HW components to aid in solving sd card interface power cycle issue.

Introducing an additional delay for those boards which had this resolved in the layout
might not be advisable, therefore I believe that target board vendors can decide for
themselves if this fix is applicable to them.

Nevertheless, it is a good point: I was planning to propose Compulab to have a look at
this patch as well to identify if this applies to their new IoT Gateway device as I saw
patches from them yesterday on the list. Was just waiting for a general feedback on the
patch itself.

> 
> >         };
> >
> >         buck2_reg: regulator-buck2 {
> > diff --git a/arch/arm/dts/imx8mq-phanbell.dts
> > b/arch/arm/dts/imx8mq-phanbell.dts
> > index 4892ad5ee1..bc6b2638ee 100644
> > --- a/arch/arm/dts/imx8mq-phanbell.dts
> > +++ b/arch/arm/dts/imx8mq-phanbell.dts
> > @@ -34,6 +34,7 @@
> >                 regulator-max-microvolt = <3300000>;
> >                 gpio = <&gpio2 19 GPIO_ACTIVE_HIGH>;
> >                 enable-active-high;
> > +               u-boot,off-on-delay-us = <20000>;
> >         };
> >  };
> >
> > --
> > 2.17.1
> >

Regards,
Andrey