Re: [PATCH] arm64: dts: rockchip: fix USB regulator on ROCK64

[Lorenz Brun <lorenz@brun.one>]

On Tue, Jan 10 2023 at 14:02:47 +01:00:00, Heiko Stübner 
<heiko@sntech.de> wrote:
> Am Donnerstag, 5. Januar 2023, 01:29:47 CET schrieb Peter Geis:
>>  On Wed, Jan 4, 2023 at 6:55 PM Lorenz Brun <lorenz@brun.one> wrote:
>>  >
>>  >
>>  > On Wed, Jan 4 2023 at 18:46:25 -05:00:00, Peter Geis
>>  > <pgwipeout@gmail.com> wrote:
>>  > > On Wed, Jan 4, 2023 at 3:55 PM Lorenz Brun <lorenz@brun.one> 
>> wrote:
>>  > >>
>>  > >>  Currently the ROCK64 device tree specifies two regulators,
>>  > >> vcc_host_5v
>>  > >>  and vcc_host1_5v for USB VBUS on the device. Both of those are
>>  > >> however
>>  > >>  specified with RK_PA2 as the GPIO enabling them, causing the
>>  > >> following
>>  > >>  error when booting:
>>  > >>
>>  > >>    rockchip-pinctrl pinctrl: pin gpio0-2 already requested by
>>  > >> vcc-host-5v-regulator; cannot claim for vcc-host1-5v-regulator
>>  > >>    rockchip-pinctrl pinctrl: pin-2 (vcc-host1-5v-regulator) 
>> status
>>  > >> -22
>>  > >>    rockchip-pinctrl pinctrl: could not request pin 2 (gpio0-2) 
>> from
>>  > >> group usb20-host-drv  on device rockchip-pinctrl
>>  > >>    reg-fixed-voltage vcc-host1-5v-regulator: Error applying 
>> setting,
>>  > >> reverse things back
>>  > >>
>>  > >>  Looking at the schematic, there are in fact three USB 
>> regulators,
>>  > >>  vcc_host_5v, vcc_host1_5v and vcc_otg_v5. But the enable 
>> signal for
>>  > >> all
>>  > >>  three is driven by Q2604 which is in turn driven by 
>> GPIO_A2/PA2.
>>  > >>
>>  > >>  Since these three regulators are not controllable separately, 
>> I
>>  > >> removed
>>  > >>  the second one which was causing the error and left a comment
>>  > >> explaining
>>  > >>  that this regulator actually controls multiple rails.
>>  > >>
>>  > >>  Signed-off-by: Lorenz Brun <lorenz@brun.one>
>>  > >>  ---
>>  > >>   arch/arm64/boot/dts/rockchip/rk3328-rock64.dts | 14 
>> +++-----------
>>  > >>   1 file changed, 3 insertions(+), 11 deletions(-)
>>  > >>
>>  > >>  diff --git a/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
>>  > >> b/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
>>  > >>  index f69a38f42d2d..bd82bc80444d 100644
>>  > >>  --- a/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
>>  > >>  +++ b/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
>>  > >>  @@ -37,6 +37,9 @@ vcc_sd: sdmmc-regulator {
>>  > >>                  vin-supply = <&vcc_io>;
>>  > >>          };
>>  > >>
>>  > >>  +       // vcc_host_5v also controls the vcc_host1_5v and
>>  > >> vcc_otg_5v rails
>>  > >>  +       // but there is only one common control signal
>>  > >> (USB20_HOST_DRV) at
>>  > >>  +       // GPIO_A2
>>  > >>          vcc_host_5v: vcc-host-5v-regulator {
>>  > >>                  compatible = "regulator-fixed";
>>  > >>                  gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_LOW>;
>>  > >>  @@ -48,17 +51,6 @@ vcc_host_5v: vcc-host-5v-regulator {
>>  > >>                  vin-supply = <&vcc_sys>;
>>  > >>          };
>>  > >>
>>  > >>  -       vcc_host1_5v: vcc_otg_5v: vcc-host1-5v-regulator {
>>  > >>  -               compatible = "regulator-fixed";
>>  > >>  -               gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_LOW>;
>>  > >>  -               pinctrl-names = "default";
>>  > >>  -               pinctrl-0 = <&usb20_host_drv>;
>>  > >>  -               regulator-name = "vcc_host1_5v";
>>  > >>  -               regulator-always-on;
>>  > >>  -               regulator-boot-on;
>>  > >>  -               vin-supply = <&vcc_sys>;
>>  > >>  -       };
>>  > >
>>  > > Fixed-regulator supports multiple regulators sharing a gpio, 
>> the issue
>>  > > is you have the pinctrl assigned multiple times which is not
>>  > > supported. Simply removing the pinctrl from all but one of the
>>  > > regulators will solve this issue.
>>  > Sure, I can just remove the pinctrl. Should I do anything about 
>> the
>>  > fact that there are three USB switches on that GPIO, but only two 
>> of
>>  > them are described as regulators here? Seems a bit inconsistent 
>> to me.
>> 
>>  All hardware *should* be described, though it isn't uncommon to see 
>> a
>>  single fixed-regulator describe several individual switches that are
>>  all fed from a common source and controlled by the same gpio. If 
>> they
>>  are not fed by a common source (for example, the otg port is often 
>> fed
>>  from a separate regulator), they must be modeled separately.
> 
> Which is essentially what Lorenz' patch already did - moving the 3 
> switches
> into one regulator. So it essentially comes down to where does the 
> not-yet
> modeled otg-regulator get its current from - also vcc_sys?
> 
Yeah, all regulators (they are just USB load/overload switches) are 
powered from VCC_SYS, there aren't that many power rails on the ROCK64.

The alternative to my first patch would essentially be this:

--- a/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
+++ b/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts
@@ -48,11 +48,18 @@ vcc_host_5v: vcc-host-5v-regulator {
                vin-supply = <&vcc_sys>;
        };

-       vcc_host1_5v: vcc_otg_5v: vcc-host1-5v-regulator {
+       vcc_host1_5v: vcc-host1-5v-regulator {
+               compatible = "regulator-fixed";
+               gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_LOW>;
+               regulator-name = "vcc_host1_5v";
+               regulator-always-on;
+               regulator-boot-on;
+               vin-supply = <&vcc_sys>;
+       };
+
+       vcc_otg_5v: vcc-otg-5v-regulator {
                compatible = "regulator-fixed";
                gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_LOW>;
-               pinctrl-names = "default";
-               pinctrl-0 = <&usb20_host_drv>;
                regulator-name = "vcc_host1_5v";
                regulator-always-on;
                regulator-boot-on;

This drops the pinctrl from all but the first regulator (which is a bit 
weird, the first one is not really special) and then describes three 
separate regulators driven by PA2.

I have no real preference, the one-regulator solution has the downside 
of not accurately describing the three separate load switches, the 
three-regulator solution has the downside of the pinctrl definitions 
being weird.

Regards,
Lorenz




_______________________________________________
Linux-rockchip mailing list
Linux-rockchip@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-rockchip