Re: [PATCH v7 7/7] arm64: dts: imx8mm: imx8mp: Add FIO1-Audio DTOs for Data Modul i.MX8M Mini and Plus eDM SBC

[Frank Li <Frank.li@oss.nxp.com>]

On Wed, Jul 29, 2026 at 10:10:32PM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 6 potential issue(s) to consider:
> - [High] Hardware Interrupt Routing Mismatch causes PCA9554 GPIO expander to conflict with CAN1 on i.MX8MP.
> - [High] The `fsl,sai-asynchronous` property breaks audio capture (and playback) on both i.MX8MM and i.MX8MP due to missing clock routing.
> - [High] The i.MX8MP overlay fails to configure the clock parent and rate for `sai5clk`, resulting in an invalid MCLK for the SGTL5000 codec.
> - [High] Duplicate `unique_unit_address` causes devres memory region conflict between `sai5clk` and `sai5` drivers.
> - [High] The shared `.dtsi` unconditionally disables `uart4`, breaking onboard Bluetooth on the i.MX8MP SBC.
> - [Low] Commit message inaccurately claims the overlay provides a "PWM Beeper" and "additional UARTs".
> --
>
> --- Patch [7]: [PATCH v7 7/7] arm64: dts: imx8mm: imx8mp: Add FIO1-Audio DTOs for Data Modul i.MX8M Mini and Plus eDM SBC ---
> commit 43246870359af0cd7d3d40f8db5eaaff8fad43a1
> Author: Marek Vasut <marex@nabladev.com>
>
> arm64: dts: imx8mm: imx8mp: Add FIO1-Audio DTOs for Data Modul i.MX8M Mini and Plus eDM SBC
>
> Add DT overlay for feature connector expansion module eDM-MOD-iMX8Mm-FIO1-Audio
> providing additional UARTs, CAN, PWM Beeper, I2C, SPI, GPIO breakout and
> SGTL5000 codec. This adapter can be optionally populated onto the eDM SBC.
>
> > Add DT overlay for feature connector expansion module eDM-MOD-iMX8Mm-FIO1-Audio
> > providing additional UARTs, CAN, PWM Beeper, I2C, SPI, GPIO breakout and
> > SGTL5000 codec.
>
> [Severity: Low]
> Is this description accurate? The commit message mentions a PWM Beeper and
> additional UARTs, but the code below disables uart4 and repurposes the
> SPDIF_TX pin used for the beeper as a GPIO instead.
>
> [ ... ]
>
> > diff --git a/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtsi b/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtsi
> > new file mode 100644
> > index 0000000000000..01cf5c89dbff4
> > --- /dev/null
> > +++ b/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtsi
> [ ... ]
> > +	gpio_feature: io-expander@20 {
> > +		compatible = "nxp,pca9554";
> > +		reg = <0x20>;
> > +		#gpio-cells = <2>;
> > +		gpio-controller;
> > +		#interrupt-cells = <2>;
> > +		interrupt-controller;
> > +		interrupt-parent = <&gpio5>;
> > +		interrupts = <4 IRQ_TYPE_LEVEL_LOW>;
>
> [Severity: High]
> Does this interrupt routing conflict on the i.MX8MP baseboard?
>
> The shared overlay hardcodes the PCA9554 interrupt to GPIO5_IO04, which is
> correct for i.MX8MM. However, on i.MX8MP, this pad (SPDIF_RX) is physically
> routed and configured for CAN1_RX.
>
> Could this cause the GPIO expander interrupts to fail and CAN traffic to
> trigger spurious GPIO interrupts when loaded on the i.MX8MP? Would it be
> better to override this interrupt in the i.MX8MP overlay?

Yes, I check flexcan1 use this pad at i.MX8MP. you need disable flexcan1
at i.mx8mp's dtso file.

>
> [ ... ]
> > +&sai2 {
> > +	pinctrl-names = "default";
> > +	pinctrl-0 = <&pinctrl_sai2>;
> > +	assigned-clock-rates = <24576000>;
> > +	fsl,sai-asynchronous;
>
> [Severity: High]
> Could this fsl,sai-asynchronous property break audio capture?
>
> It forces the SAI block to use independent clocks for TX and RX. However, the
> pinmux routing on i.MX8MM only provides RX_BCLK and TX_SYNC, meaning RX
> lacks a SYNC clock. On i.MX8MP, only TX clocks are routed, leaving RX without
> both BCLK and SYNC clocks.
>
> [ ... ]
> > +/* UART4 is blocked by RDC and used as CM4 console UART */
> > +&uart4 {	/* UART to 1-Wire J5 */
> > +	status = "disabled";
> > +};
>
> [Severity: High]
> Will disabling uart4 here break the onboard Bluetooth on the i.MX8MP SBC?
>
> This file is shared between the i.MX8MM and i.MX8MP overlays, but on the
> i.MX8MP SBC, uart4 is actively used for the infineon,cyw55572-bt module.
>
> [ ... ]
> > diff --git a/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso b/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso
> > new file mode 100644
> > index 0000000000000..4a02bd026ac16
> > --- /dev/null
> > +++ b/arch/arm64/boot/dts/freescale/imx8mm-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso
> [ ... ]
> > +	sai5clk: clock-controller@30050000 {	/* SAI5 */
> > +		compatible = "fsl,imx8mm-sai-clock", "fsl,imx8mq-sai-clock";
> > +		reg = <0x30050000 0x10000>;
>
> [Severity: High]
> Can this cause a devres memory region conflict during driver probe?
>
> This overlay adds a new clock-controller node with the exact same unit address
> and register range as the existing sai@30050000 audio node.
>
> Since both drivers will use devm_platform_ioremap_resource() to request the
> region exclusively, whichever probes second might receive -EBUSY and fail.

at imx8mm.dts, 30050000 is SAI5, here you change to to use clock-controller

Suppose address here is wrong.

Frank

>
> [ ... ]
> > diff --git a/arch/arm64/boot/dts/freescale/imx8mp-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso b/arch/arm64/boot/dts/freescale/imx8mp-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso
> > new file mode 100644
> > index 0000000000000..90c2dd1e67248
> > --- /dev/null
> > +++ b/arch/arm64/boot/dts/freescale/imx8mp-data-modul-edm-sbc-overlay-edm-mod-imx8mm-fio1-audio.dtso
> [ ... ]
> > +	sai5clk: clock-controller@30c50000 {
> > +		compatible = "fsl,imx8mp-sai-clock", "fsl,imx8mq-sai-clock";
> > +		reg = <0x30c50000 0x10000>;
>
> [Severity: High]
> Similar to the i.MX8MM overlay, can this cause a memory region conflict with
> the existing sai@30c50000 audio node?
>
> [ ... ]
> > +		#clock-cells = <1>;
> > +		clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_SAI5_IPG>,
> > +			 <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_SAI5_MCLK1>;
> > +		clock-names = "bus", "mclk1";
> > +		pinctrl-names = "default";
> > +		pinctrl-0 = <&pinctrl_codec_mclk>;
> > +		status = "okay";
> > +	};
> > +};
>
> [Severity: High]
> Does this node also need assigned-clocks and assigned-clock-rates configured?
>
> Unlike the i.MX8MM variant, this node leaves the MCLK unconfigured. The
> SGTL5000 driver and simple-audio-card do not dynamically set this clock rate
> via clk_set_rate() unless mclk-fs is defined.
>
> Could this leave the clock at 0 Hz and prevent the codec from functioning on
> the i.MX8MP?
>
> --
> Sashiko AI review · https://sashiko.dev/#/patchset/20260729214723.557461-1-marex@nabladev.com?part=7