by Alistair Francis

[permalink] [raw]

Subject: Re: [PATCH v2 3/3] ARM: dts: imx7d-remarkable2: Enable the rohm,bd71815

On Sat, Jan 21, 2023 at 12:01 AM Matti Vaittinen
<[email protected]> wrote:
>
> On 1/20/23 11:17, Alistair Francis wrote:
> > On Tue, Jan 10, 2023 at 5:13 AM Matti Vaittinen
> > <[email protected]> wrote:
> >>
> >> ma 9. tammik. 2023 klo 14.32 Alistair Francis ([email protected]) kirjoitti:
> >>>
> >>> On Tue, Dec 20, 2022 at 11:17 PM Marco Felsch <[email protected]> wrote:
> >>>>
> >>>> Hi Alistair,
> >>>>
> >>>> thanks for your patch, please see below.
> >>>>
> >>>> On 22-12-20, Alistair Francis wrote:
> >>>>> Add support for the rohm,bd71815 power controller controller for the
> >>>>> reMarkable 2.
> >>>>>
> >>>>> Signed-off-by: Alistair Francis <[email protected]>
> >>>>> ---
> >>>>> arch/arm/boot/dts/imx7d-remarkable2.dts | 159 ++++++++++++++++++++++++
> >>>>> 1 file changed, 159 insertions(+)
> >>>>>
> >>>>> diff --git a/arch/arm/boot/dts/imx7d-remarkable2.dts b/arch/arm/boot/dts/imx7d-remarkable2.dts
> >>>>> index a138b292ec6a..4387d30d6180 100644
> >>>>> --- a/arch/arm/boot/dts/imx7d-remarkable2.dts
> >>>>> +++ b/arch/arm/boot/dts/imx7d-remarkable2.dts
> >>>>> @@ -92,6 +92,10 @@ wifi_pwrseq: wifi_pwrseq {
> >>>>> };
> >>>>> };
> >>>>>
> >>>>> +&cpu0 {
> >>>>> + cpu-supply = <&buck1_reg>;
> >>>>> +};
> >>>>> +
> >>>>> &clks {
> >>>>> assigned-clocks = <&clks IMX7D_CLKO2_ROOT_SRC>,
> >>>>> <&clks IMX7D_CLKO2_ROOT_DIV>;
> >>>>> @@ -119,6 +123,148 @@ wacom_digitizer: digitizer@9 {
> >>>>> };
> >>>>> };
> >>>>>
> >>>>> +&i2c2 {
> >>>>> + clock-frequency = <100000>;
> >>>>
> >>>> We can set this now to 400kHz since the driver can handle quirks now
> >>>> internal.
> >>>
> >>> After a discussion on the first patch I'm going to leave this as is.
> >>> Let me know if you do want me to change it.
> >>>
> >>>>
> >>>>> + pinctrl-names = "default", "sleep";
> >>>>> + pinctrl-0 = <&pinctrl_i2c2>;
> >>>>> + pinctrl-1 = <&pinctrl_i2c2>;
> >>>>
> >>>> Do we need a "sleep" config here since this will be the same
> >>>> configuration as the "default" one.
> >>>
> >>> Nope, dropped.
> >>>
> >>>>
> >>>>> + status = "okay";
> >>>>> +
> >>>>> + bd71815: pmic@4b {
> >>>>> + compatible = "rohm,bd71815";
> >>>>> + reg = <0x4b>;
> >>>>> + pinctrl-names = "default";
> >>>>> + pinctrl-0 = <&pinctrl_bd71815>;
> >>>>> + interrupt-parent = <&gpio6>; /* PMIC_INT_B GPIO6_IO16 */
> >>>>> + interrupts = <16 IRQ_TYPE_LEVEL_LOW>;
> >>>>> + gpio-controller;
> >>>>> + clocks = <&clks IMX7D_CLKO2_ROOT_SRC>;
> >>>>> + clock-output-names = "bd71815-32k-out";
> >>>>> + #clock-cells = <0>;
> >>>>> + #gpio-cells = <1>;
> >>>>> +
> >>>>> + regulators {
> >>>>> + #address-cells = <1>;
> >>>>> + #size-cells = <0>;
> >>>>> +
> >>>>> + buck1_reg: regulator@0 {
> >>>>> + reg = <0>;
> >>>>> + regulator-compatible = "buck1";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <2000000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + regulator-ramp-delay = <1250>;
> >>>>> + };
> >>>>> +
> >>>>> + buck2_reg: regulator@1 {
> >>>>> + reg = <1>;
> >>>>> + regulator-compatible = "buck2";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <2000000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + regulator-ramp-delay = <1250>;
> >>>>> + };
> >>>>> +
> >>>>> + buck3_reg: regulator@2 {
> >>>>> + reg = <2>;
> >>>>> + regulator-compatible = "buck3";
> >>>>> + regulator-min-microvolt = <1200000>;
> >>>>> + regulator-max-microvolt = <2700000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + buck4_reg: regulator@3 {
> >>>>> + reg = <3>;
> >>>>> + regulator-compatible = "buck4";
> >>>>> + regulator-min-microvolt = <1100000>;
> >>>>> + regulator-max-microvolt = <1850000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + buck5_reg: regulator@4 {
> >>>>> + reg = <4>;
> >>>>> + regulator-compatible = "buck5";
> >>>>> + regulator-min-microvolt = <1800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + ldo1_reg: regulator@5 {
> >>>>> + reg = <5>;
> >>>>> + regulator-compatible = "ldo1";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + ldo2_reg: regulator@6 {
> >>>>> + reg = <6>;
> >>>>> + regulator-compatible = "ldo2";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + ldo3_reg: regulator@7 {
> >>>>> + reg = <7>;
> >>>>> + regulator-compatible = "ldo3";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + ldo4_reg: regulator@8 {
> >>>>> + reg = <8>;
> >>>>> + regulator-compatible = "ldo4";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + ldo5_reg: regulator@9 {
> >>>>> + reg = <9>;
> >>>>> + regulator-compatible = "ldo5";
> >>>>> + regulator-min-microvolt = <800000>;
> >>>>> + regulator-max-microvolt = <3300000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + dvref_reg: regulator@a {
> >>>>> + reg = <0xa>;
> >>>>> + regulator-compatible = "dvref";
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + lpsr_reg: regulator@b {
> >>>>> + reg = <0xb>;
> >>>>> + regulator-compatible = "lpsr";
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>> +
> >>>>> + wled_reg: regulator@c {
> >>>>> + reg = <0xc>;
> >>>>> + regulator-compatible = "wled";
> >>>>> + regulator-min-microamp = <10>;
> >>>>> + regulator-max-microamp = <25000>;
> >>>>> + regulator-boot-on;
> >>>>> + regulator-always-on;
> >>>>> + };
> >>>>
> >>>> Note: You have marked all regulators as always on, this is rather
> >>>> suboptimal due to power consumption.
> >>>
> >>> Good point. The power consumption is already pretty bad, as mainline
> >>> doesn't support the i.MX deep sleep. As I know this works I'd like to
> >>> keep it as is. I'll test to see if removing them doesn't break
> >>> anything and then send a follow up patch.
> >>
> >> I'd like to suggest also seeing what removing the "regulator-boot-on"
> >> causes. As far as I understand, the "regulator-boot-on" mainly (only?)
> >> intended to be used in situations where the regulator state can not be
> >> reliably read from the hardware at the boot-up. AFAIR, unlike the
> >> BD71837 and BD71847,
> >> the BD71815 does not have this limitation. (This is just my
> >> understanding - which is based on some age-old discussions with Mark
> >> Brown - hence this comment is just a suggestion to try this while
> >> testing, not a comment intended to be demanding for a change)
> >
> > I have tried removing both "regulator-boot-on" and "regulator-always-on".
> >
> > After one boot to userspace (where the display didn't work correctly)
> > the device now seems to not power up at all. Even u-boot is failing to
> > start. I'm not sure if the regulators are now in some strange state or
> > if something is fried.
>
> Ouch. This does not sound good at all. Sorry for a bad advice. :(

Yeah, it was a bit of a shock! No worries, the advice was good! Just
unlucky that the vendor hardware and kernel aren't up to par.

>
> I can only assume the "power dependencies" are not modelled correctly -
> Eg. some required power-rails are not marked to be used by devices that
> actually need them. This is probably being worked around using the
> regulator-boot-on / regulator-always-on - which will cause the Linux to
> turn on these rails when the driver is load. My understanding is that
> this is not the intended use-case for the regulator-boot-on flag though.

Yeah, I'm guessing that's what is happening. It seems that somehow a
rail for the power controller is being turned off. As the device can't
be woken up with the physical button or a charger about 30 seconds
after boot.

> Also, as was said - the side-effect of always-on is that there will be
> no power-saving options - which is still better for an user than a fried
> device would be :|

I think for now that's what we are stuck with. It's very difficult to
debug as the device is sealed.

>
> >
> > I'm currently waiting for the battery to discharge to see if that helps.
> >
>
> /me keeps fingers crossed. You wouldn't guess how hard it was to type this!

Thanks! Once the battery depleted I managed to get signs of life. Now
I just need to quickly revert to the old device tree :)

Alistair

>
> Yours,
> -- Matti
>
> --
> Matti Vaittinen
> Linux kernel developer at ROHM Semiconductors
> Oulu Finland
>
> ~~ When things go utterly wrong vim users can always type :help! ~~
>