Re: [PATCH v7 1/1] arm64: dts: add tqma9596la-mba95xxca

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

On Tue, Jul 14, 2026 at 02:22:54PM +0200, Alexander Stein wrote:
> Hi Frank,
>
> Am Montag, 13. Juli 2026, 18:14:50 CEST schrieb Frank Li:
> > On Tue, Jul 07, 2026 at 06:59:55AM +0000, sashiko-bot@kernel.org wrote:
> > > Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
> > > - [High] Conflicting GPIO assignments for M.2 Key-E SDIO Reset between the `wifi-connector` device node and a leftover `gpio-hog`.
> > > - [High] The fan supply regulator `reg_pwm_fan` is never enabled, leaving the fan unpowered.
> > > --
> > >
> > > --- Patch [1]: [PATCH v7 1/1] arm64: dts: add tqma9596la-mba95xxca ---
> > > commit c06cc8eb19eb4a69fc0547caa6773b503beae3a7
> > > Author: Markus Niebel <Markus.Niebel@ew.tq-group.com>
> > >
> > > arm64: dts: add tqma9596la-mba95xxca
> > >
> > > This patch adds support for the TQMa95xxLA modules on the MBa95xxCA carrier
> > > board. It introduces a common device tree include for the module and a specific
> > > device tree for the carrier board.
> > >
> > > > diff --git a/arch/arm64/boot/dts/freescale/imx95-tqma9596la-mba95xxca.dts b/arch/arm64/boot/dts/freescale/imx95-tqma9596la-mba95xxca.dts
> > > > new file mode 100644
> > > > index 0000000000000..db2fd6991d112
> > > > --- /dev/null
> > > > +++ b/arch/arm64/boot/dts/freescale/imx95-tqma9596la-mba95xxca.dts
> > > [ ... ]
> > > > +	wifi-connector {
> > > > +		compatible = "pcie-m2-e-connector";
> > > > +		vpcie3v3-supply = <&reg_v3v3_mb>;
> > > > +		vpcie1v8-supply = <&reg_v1v8_mb>;
> > > > +		i2c-parent = <&lpi2c2>;
> > > > +		w-disable1-gpios = <&expander2 9 GPIO_ACTIVE_LOW>;
> > > > +		w-disable2-gpios = <&expander2 10 GPIO_ACTIVE_LOW>;
> > > > +		/* UART_WAKE is connected to SM */
> > > > +		/* SDIO_WAKE is connected to SM */
> > > > +		sdio-reset-gpios = <&expander3 2 GPIO_ACTIVE_LOW>;
> > > > +	};
> > >
> > > [Severity: High]
> > > Does this sdio-reset-gpios assignment conflict with the gpio-hog on the same
> > > line?
> > >
> > > The wifi-connector node requests expander3 line 2 here, but later in the file,
> > > m2-keye-sdio-rst-hog hogs the exact same GPIO line. Won't this conflict cause
> > > the wifi-connector driver to fail probing with -EBUSY?
> >
> > suppose needn't gpio hog for expander3 2 ?
> > can you double check it?
>
> Well pwrseq-pcie-m2.c doesn't support "sdio-reset-gpios" yet, so there is no -EBUSY.
> But I'm okay with remove the hog which seems appropriate.
>
> > >
> > > [ ... ]
> > > > +	fan_controller: fan-controller@2f {
> > > > +		compatible = "microchip,emc2301", "microchip,emc2305";
> > > > +		reg = <0x2f>;
> > > > +		#address-cells = <1>;
> > > > +		#size-cells = <0>;
> > > > +		#pwm-cells = <3>;
> > > > +		status = "okay";
> > > > +
> > > > +		fan: fan@0 {
> > > > +			reg = <0x0>;
> > > > +			pwms = <&fan_controller 40000 PWM_POLARITY_INVERTED 1>;
> > > > +			#cooling-cells = <2>;
> > > > +			fan-supply = <&reg_pwm_fan>;
> > > > +		};
> > > > +	};
> > >
> > > [Severity: High]
> > > Will the reg_pwm_fan regulator ever be enabled?
> > >
> > > The emc2305 driver does not parse or request the fan-supply property from
> > > child nodes. Since reg_pwm_fan is not marked regulator-always-on earlier in
> > > this file, does this leave the fan unpowered?
> >
> > look like make sense, but dts itself should be correct. You'd better update
> > driver also.
>
> I'ts on my TODO, but I won't manage to add it for the time being. How do you
> want to procede? I'd rather remove the fan-supply and disable the reg_pwm_fan
> for the time being.

leave reg_pwm_fan always on and add comments, after fan support it, we
can remove always on

Frank

>
> Best regards,
> Alexander
> --
> TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
> Amtsgericht München, HRB 105018
> Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
> http://www.tq-group.com/