Re: [PATCH v1 1/1] arm64: dts: imx8mp-debix-model-a: Add HDMI output support

From: Christopher Obbard
Date: Mon Apr 15 2024 - 13:10:50 EST


Hi Laurent,

On Mon, 2024-04-15 at 19:35 +0300, Laurent Pinchart wrote:
> On Mon, Apr 15, 2024 at 04:08:10PM +0100, Kieran Bingham wrote:
> > Hi Chris,
> >
> >
> > Quoting Christopher Obbard (2024-04-15 12:41:27)
> > > Enable the HDMI output on the Debix Model A SBC, using the HDMI encoder
> > > present in the i.MX8MP SoC.
> >
> > Aha, you beat me to it. I have a commit locally (Dated 2022-09-06) but
> > not sent because I didn't realise the HDMI support finally got upstream
> > \o/
> >
> > > Signed-off-by: Christopher Obbard <chris.obbard@xxxxxxxxxxxxx>
> > > ---
> > >
> > >  .../dts/freescale/imx8mp-debix-model-a.dts    | 47 +++++++++++++++++++
> > >  1 file changed, 47 insertions(+)
> > >
> > > diff --git a/arch/arm64/boot/dts/freescale/imx8mp-debix-model-a.dts
> > > b/arch/arm64/boot/dts/freescale/imx8mp-debix-model-a.dts
> > > index 2c19766ebf09..29529c2ecac9 100644
> > > --- a/arch/arm64/boot/dts/freescale/imx8mp-debix-model-a.dts
> > > +++ b/arch/arm64/boot/dts/freescale/imx8mp-debix-model-a.dts
> > > @@ -20,6 +20,18 @@ chosen {
> > >                 stdout-path = &uart2;
> > >         };
> > >  
> > > +       hdmi-connector {
> > > +               compatible = "hdmi-connector";
> > > +               label = "hdmi";
> > > +               type = "a";
> > > +
> > > +               port {
> > > +                       hdmi_connector_in: endpoint {
> > > +                               remote-endpoint = <&hdmi_tx_out>;
> > > +                       };
> > > +               };
> > > +       };
> > > +
> >
> > Interesting. My patch missed this. But it looks correct.
> >
> > >         leds {
> > >                 compatible = "gpio-leds";
> > >                 pinctrl-names = "default";
> > > @@ -94,6 +106,28 @@ ethphy0: ethernet-phy@0 { /* RTL8211E */
> > >         };
> > >  };
> > >  
> > > +&hdmi_pvi {
> > > +       status = "okay";
> > > +};
> > > +
> > > +&hdmi_tx {
> > > +       pinctrl-names = "default";
> > > +       pinctrl-0 = <&pinctrl_hdmi>;
> > > +       status = "okay";
> > > +
> > > +       ports {
> > > +               port@1 {
> > > +                       hdmi_tx_out: endpoint {
> > > +                               remote-endpoint = <&hdmi_connector_in>;
> > > +                       };
> > > +               };
> > > +       };
> > > +};
> > > +
> > > +&hdmi_tx_phy {
> > > +       status = "okay";
> > > +};
> > > +
> > >  &i2c1 {
> > >         clock-frequency = <400000>;
> > >         pinctrl-names = "default";
> > > @@ -241,6 +275,10 @@ &i2c6 {
> > >         status = "okay";
> > >  };
> > >  
> > > +&lcdif3 {
> > > +       status = "okay";
> > > +};
> > > +
> >
> > Except for the addition of the connector, the above matches my patch to
> > here.
> >
> > >  &snvs_pwrkey {
> > >         status = "okay";
> > >  };
> >
> > But in my patch I have the following hunk here: (I haven't checked to
> > see if this still applies on mainline, so take with a pinch of salt if
> > it's not there!)
> >
> >
> >  &iomuxc {
> >   pinctrl-names = "default";
> > - pinctrl-0 = <&pinctrl_hog>;
> > -
> > - pinctrl_hog: hoggrp {
> > - fsl,pins = <
> > -
> > MX8MP_IOMUXC_HDMI_DDC_SCL__HDMIMIX_HDMI_SCL 0x400001c3
> > -
> > MX8MP_IOMUXC_HDMI_DDC_SDA__HDMIMIX_HDMI_SDA 0x400001c3
> > -
> > MX8MP_IOMUXC_HDMI_HPD__HDMIMIX_HDMI_HPD 0x40000019
> > -
> > MX8MP_IOMUXC_HDMI_CEC__HDMIMIX_HDMI_CEC 0x40000019
> > - >;
> > - };
> >
> >   pinctrl_eqos: eqosgrp {
> >   fsl,pins = <
> >   MX8MP_IOMUXC_ENET_MDC__ENET_QOS_MDC
> > 0x3
> >   MX8MP_IOMUXC_ENET_MDIO__ENET_QOS_MDIO
> > 0x3
> >
> >
> > > @@ -358,6 +396,15 @@
> > > MX8MP_IOMUXC_NAND_READY_B__GPIO3_IO16                              0x19
> > >                 >;
> > >         };
> > >  
> > > +       pinctrl_hdmi: hdmigrp {
> > > +               fsl,pins = <
> > > +                      
> > > MX8MP_IOMUXC_HDMI_DDC_SCL__HDMIMIX_HDMI_SCL                    
> > > 0x400001c3
> > > +                      
> > > MX8MP_IOMUXC_HDMI_DDC_SDA__HDMIMIX_HDMI_SDA                    
> > > 0x400001c3
> > > +                      
> > > MX8MP_IOMUXC_HDMI_HPD__HDMIMIX_HDMI_HPD                        
> > > 0x40000019
> > > +                      
> > > MX8MP_IOMUXC_HDMI_CEC__HDMIMIX_HDMI_CEC                        
> > > 0x40000019
> > > +               >;
> > > +       };
> > > +
> >
> > And my addition here is :
> >
> >
> > + pinctrl_hdmi: hdmigrp {
> > + fsl,pins = <
> > + MX8MP_IOMUXC_HDMI_DDC_SCL__HDMIMIX_HDMI_SCL 0
> > x1c3
> > + MX8MP_IOMUXC_HDMI_DDC_SDA__HDMIMIX_HDMI_SDA 0
> > x1c3
> > + MX8MP_IOMUXC_HDMI_HPD__HDMIMIX_HDMI_HPD
> > 0x19
> > + MX8MP_IOMUXC_HDMI_CEC__HDMIMIX_HDMI_CEC
> > 0x19
> > + >;
> > + };
> > +
> >
> >
> > I haven't looked into what the 0x40000000 does yet, but just
> > highlighting the difference from the version I've been using to make use
> > of HDMI so far.
> >
> > Does anyone else know the impact here? Otherwise I'll try to find time
> > to check this later. (For some undefined term of later...)
>
> In drivers/pinctrl/freescale/pinctrl-imx.c,
>
> #define IMX_NO_PAD_CTL  0x80000000      /* no pin config need */
> #define IMX_PAD_SION 0x40000000         /* set SION */
>
> The SION (Software Input ON) bit forces the input path active for the
> pin. This can be used, for instance, to capture through GPIO the value
> of a pin driven by a module. I'm not sure that's needed here.

Thanks for the explanation, makes perfect sense. I will send a v2 without the
SION bit set (e.g exactly per the hunk in Kieran's patch).


Chris

>
> > >         pinctrl_i2c1: i2c1grp {
> > >                 fsl,pins = <
> > >                        
> > > MX8MP_IOMUXC_I2C1_SCL__I2C1_SCL                                
> > > 0x400001c2
>