Re: [PATCH v6 1/2] arm64: dts: qcom: sc7280: Add pinmux for I2S speaker and Headset

From: Matthias Kaehlcke
Date: Tue Apr 05 2022 - 16:41:23 EST


On Tue, Apr 05, 2022 at 04:42:46PM +0530, Srinivasa Rao Mandadapu wrote:
> Add AMP enable node and pinmux for primary and secondary I2S
> for SC7280 based platforms.
>
> Signed-off-by: Srinivasa Rao Mandadapu <quic_srivasam@xxxxxxxxxxx>
> Co-developed-by: Venkata Prasad Potturu <quic_potturu@xxxxxxxxxxx>
> Signed-off-by: Venkata Prasad Potturu <quic_potturu@xxxxxxxxxxx>
> ---
> arch/arm64/boot/dts/qcom/sc7280-herobrine.dtsi | 34 +++++++++++++++++++++
> arch/arm64/boot/dts/qcom/sc7280-idp.dtsi | 20 +++++++++++++
> arch/arm64/boot/dts/qcom/sc7280.dtsi | 41 ++++++++++++++++++++++++++
> 3 files changed, 95 insertions(+)
>
> diff --git a/arch/arm64/boot/dts/qcom/sc7280-herobrine.dtsi b/arch/arm64/boot/dts/qcom/sc7280-herobrine.dtsi
> index dc17f20..de646d9 100644
> --- a/arch/arm64/boot/dts/qcom/sc7280-herobrine.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sc7280-herobrine.dtsi
> @@ -530,6 +530,26 @@ ap_ec_spi: &spi10 {
> drive-strength = <2>;
> };
>
> +&pri_mi2s_data0 {
> + drive-strength = <6>;

Isn't this pin used as an input (HP_DIN)? Is specifying the drive strength
really needed?

> +};
> +
> +&pri_mi2s_data1 {
> + drive-strength = <6>;
> +};
> +
> +&pri_mi2s_mclk {
> + drive-strength = <6>;
> +};
> +
> +&pri_mi2s_sclk {
> + drive-strength = <6>;
> +};
> +
> +&pri_mi2s_ws {
> + drive-strength = <6>;
> +};
> +
> &qspi_cs0 {
> bias-disable;
> drive-strength = <8>;
> @@ -610,6 +630,20 @@ ap_ec_spi: &spi10 {
> drive-strength = <10>;
> };
>
> +&sec_mi2s_data0 {
> + drive-strength = <6>;
> + bias-disable;
> +};
> +
> +&sec_mi2s_sclk {
> + drive-strength = <6>;
> + bias-disable;
> +};
> +
> +&sec_mi2s_ws {
> + drive-strength = <6>;
> +};

Actually there are several sound configs for herobrine boards. For now I
think it's ok to specify the config for herobrine -rev1 (as this patch
does) and we can sort out later how to best support the different configs.

> /* PINCTRL - board-specific pinctrl */
>
> &pm7325_gpios {
> diff --git a/arch/arm64/boot/dts/qcom/sc7280-idp.dtsi b/arch/arm64/boot/dts/qcom/sc7280-idp.dtsi
> index ecbf2b8..2afbbe3 100644
> --- a/arch/arm64/boot/dts/qcom/sc7280-idp.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sc7280-idp.dtsi
> @@ -462,7 +462,27 @@
> drive-strength = <10>;
> };
>
> +&sec_mi2s_data0 {
> + drive-strength = <6>;
> + bias-disable;
> +};
> +
> +&sec_mi2s_sclk {
> + drive-strength = <6>;
> + bias-disable;
> +};
> +
> +&sec_mi2s_ws {
> + drive-strength = <6>;
> +};
> +
> &tlmm {
> + amp_en: amp-en {
> + pins = "gpio63";
> + bias-pull-down;
> + drive-strength = <2>;
> + };

nit: all the other pins are i2s related, it might make sense to add amp_en
in a separate patch.

> +
> bt_en: bt-en {
> pins = "gpio85";
> function = "gpio";
> diff --git a/arch/arm64/boot/dts/qcom/sc7280.dtsi b/arch/arm64/boot/dts/qcom/sc7280.dtsi
> index f0b64be..8d8cec5 100644
> --- a/arch/arm64/boot/dts/qcom/sc7280.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sc7280.dtsi
> @@ -3527,6 +3527,31 @@
> function = "pcie1_clkreqn";
> };
>
> + pri_mi2s_data0: pri-mi2s-data0 {
> + pins = "gpio98";
> + function = "mi2s0_data0";
> + };
> +
> + pri_mi2s_data1: pri-mi2s-data1 {
> + pins = "gpio99";
> + function = "mi2s0_data1";
> + };
> +
> + pri_mi2s_mclk: pri-mi2s-mclk {
> + pins = "gpio96";
> + function = "pri_mi2s";
> + };
> +
> + pri_mi2s_sclk: pri-mi2s-sclk {
> + pins = "gpio97";
> + function = "mi2s0_sck";
> + };
> +
> + pri_mi2s_ws: pri-mi2s-ws {
> + pins = "gpio100";
> + function = "mi2s0_ws";
> + };
> +
> qspi_clk: qspi-clk {
> pins = "gpio14";
> function = "qspi_clk";
> @@ -4261,6 +4286,22 @@
> drive-strength = <2>;
> bias-bus-hold;
> };
> +
> + sec_mi2s_data0: sec-mi2s-data0 {
> + pins = "gpio107";
> + function = "mi2s1_data0";
> + };
> +
> + sec_mi2s_sclk: sec-mi2s-sclk {
> + pins = "gpio106";
> + function = "mi2s1_sck";
> + };
> +
> + sec_mi2s_ws: sec-mi2s-ws {
> + pins = "gpio108";
> + function = "mi2s1_ws";
> + };

Is there a particular reason for the pri/sec nomenclature? The datasheet and
schematics call the pin mi2sN_xyz, it seems it would be clearer to follow
that naming. Primary/secondary seems to imply a 'master/slave' topology, but
these are independent controllers IIUC. The datasheet refers to pin 96 as
PRI_MI2S_MCLK and pin 105 SEC_MI2S_MCLK, I guess the naming was derived from
that.

My suggestion would be to follow the naming in the datasheet/schematic, i.e.
mi2sN_data0, mi2sN_data1, pri/sec_mi2s_mclk, mi2sN_sck, mi2sN_ws.