linux-kernel - Re: [PATCH 5/6] arm64: dts: ls1028a-qds: make the QIXIS CPLD use the simple-mfd-i2c.c driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <z7knkmi7kjhwlqwokikozos7whwaj5vmseh7kjrdl2la3kmj5e@wwzlsxpq573v>
Date: Wed, 7 May 2025 15:28:18 +0300
From: Ioana Ciornei <ioana.ciornei@....com>
To: Krzysztof Kozlowski <krzk@...nel.org>
Cc: Lee Jones <lee@...nel.org>, Rob Herring <robh@...nel.org>, 
	Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>, 
	Shawn Guo <shawnguo@...nel.org>, devicetree@...r.kernel.org, linux-kernel@...r.kernel.org, 
	linux-arm-kernel@...ts.infradead.org, Vladimir Oltean <vladimir.oltean@....com>
Subject: Re: [PATCH 5/6] arm64: dts: ls1028a-qds: make the QIXIS CPLD use the
 simple-mfd-i2c.c driver

On Wed, May 07, 2025 at 06:54:38AM +0200, Krzysztof Kozlowski wrote:
> On 06/05/2025 16:21, Ioana Ciornei wrote:
> > On Fri, May 02, 2025 at 09:04:03AM +0200, Krzysztof Kozlowski wrote:
> >> On Wed, Apr 30, 2025 at 06:36:33PM GMT, Ioana Ciornei wrote:
> >>> From: Vladimir Oltean <vladimir.oltean@....com>
> >>>
> >>> The MDIO mux on the LS1028A-QDS never worked in mainline. The device
> >>> tree was submitted as-is, and there is a downstream driver for the QIXIS
> >>> FPGA:
> >>>
> >>> https://github.com/nxp-qoriq/linux/blob/lf-6.12.y/drivers/soc/fsl/qixis_ctrl.c
> >>>
> >>> That driver is very similar to the already existing drivers/mfd/simple-mfd-i2c.c,
> >>> and the hardware works with the simple-mfd-i2c driver, so there isn't
> >>> any reason to upstream the other one.
> >>>
> >>> Adapt the compatible string and child node format of the FPGA node, so
> >>> that the simple-mfd-i2c driver accepts it.
> >>
> >> Why do you break the users based on some driver differences? Fix the
> >> drivers, not the DTS.
> >>
> >>>
> >>> Signed-off-by: Vladimir Oltean <vladimir.oltean@....com>
> >>> Signed-off-by: Ioana Ciornei <ioana.ciornei@....com>
> >>> ---
> >>>  arch/arm64/boot/dts/freescale/fsl-ls1028a-qds.dts | 9 +++++----
> >>>  1 file changed, 5 insertions(+), 4 deletions(-)
> >>>
> >>> diff --git a/arch/arm64/boot/dts/freescale/fsl-ls1028a-qds.dts b/arch/arm64/boot/dts/freescale/fsl-ls1028a-qds.dts
> >>> index 0bb2f28a0441..58b54d521d75 100644
> >>> --- a/arch/arm64/boot/dts/freescale/fsl-ls1028a-qds.dts
> >>> +++ b/arch/arm64/boot/dts/freescale/fsl-ls1028a-qds.dts
> >>> @@ -338,17 +338,18 @@ sgtl5000: audio-codec@a {
> >>>  	};
> >>>  
> >>>  	fpga@66 {
> >>> -		compatible = "fsl,ls1028aqds-fpga", "fsl,fpga-qixis-i2c",
> >>> -			     "simple-mfd";
> >>> +		compatible = "fsl,ls1028a-qds-qixis-i2c";
> >>
> >> This breaks all the existing users. NAK.
> > 
> > Using a mainline kernel, this DT node was never used or probed by a
> > driver since that driver was never submitted. I am not breaking any user
> > of the mainline kernel.
> 1. Users of DTS is plural, so what about all other projects and out of
> tree users?
> 2. Did you remove simple-mfd from kernel or what? How can you claim
> there is no driver for simple-mfd?

No, I did not remove simple-mfd from the kernel.

What I am saying is that using a clean linux-next tag the child devices
of this node are not probed.

For example, the LS1028AQDS's DT looks like this in linux-next:

        fpga@66 {
                compatible = "fsl,ls1028aqds-fpga", "fsl,fpga-qixis-i2c",
                             "simple-mfd";
                reg = <0x66>;

                mux: mux-controller {
                        compatible = "reg-mux";
                        #mux-control-cells = <1>;
                        mux-reg-masks = <0x54 0xf0>; /* 0: reg 0x54, bits 7:4 */
                };
        };

        (...)

        mdio-mux {
                compatible = "mdio-mux-multiplexer";
                mux-controls = <&mux 0>;
                mdio-parent-bus = <&enetc_mdio_pf3>;
                #address-cells = <1>;
                #size-cells = <0>;

                /* on-board RGMII PHY */
                mdio@0 {
                        #address-cells = <1>;
                        #size-cells = <0>;
                        reg = <0>;

                        qds_phy1: ethernet-phy@5 {
                                /* Atheros 8035 */
                                reg = <5>;
                        };
                };

                (...)
        };

        &enetc_port1 {
                phy-handle = <&qds_phy1>;
                phy-mode = "rgmii-id";
                status = "okay";
        };

If, as you say, this works just by having the simple-mfd, I should have
been able to see the enetc_port1 functional and the RGMII PHY be
accesible on the MDIO bus which is behind the reg-mux. But this is not
happening.

Instead, I get this:

	[   17.635216] platform mdio-mux: deferred probe pending: mdio-mux-multiplexer: Failed to get mux

	root@...alhost:~# ip link set dev eno1 up
	[ 1155.190391] net eno1: could not attach to PHY
	root@...alhost:~# uname -a
	Linux localhost 6.15.0-rc5-next-20250507 #112 SMP PREEMPT Wed May  7 15:21:14 EEST 2025 aarch64 aarch64 aarch64 GNU/Linux

>From what I understand, even though the fpga@66 has the simple-mfd
compatible, no entity initializes an I2C regmap (since this is an I2C
device) for it so that it can be used by any child device.

Please let me know what I am missing.

Ioana