linux-kernel - Re: [PATCH v2 1/3] mips: dts: ralink: mt7621: associate uart1

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]

Date: Fri, 8 Mar 2024 16:24:28 +0300
From: Arınç ÜNAL <arinc.unal@...nc9.com>
To: Justin Swartz <justin.swartz@...ingedge.co.za>,
 AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
Cc: Sergio Paracuellos <sergio.paracuellos@...il.com>,
 Rob Herring <robh+dt@...nel.org>,
 Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
 Conor Dooley <conor+dt@...nel.org>,
 Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
 Matthias Brugger <matthias.bgg@...il.com>, linux-mips@...r.kernel.org,
 devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
 linux-arm-kernel@...ts.infradead.org, linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH v2 1/3] mips: dts: ralink: mt7621: associate uart1_pins
 with serial0

On 8.03.2024 15:40, Justin Swartz wrote:
> Hi Angelo
> 
> On 2024-03-08 10:41, AngeloGioacchino Del Regno wrote:
>> Il 07/03/24 20:04, Justin Swartz ha scritto:
>>> Add missing pinctrl-name and pinctrl-0 properties to declare
>>> that the uart1_pins group is associated with serial0.
>>>
>>> Signed-off-by: Justin Swartz <justin.swartz@...ingedge.co.za>
>>> ---
>>>   arch/mips/boot/dts/ralink/mt7621.dtsi | 3 +++
>>>   1 file changed, 3 insertions(+)
>>>
>>> diff --git a/arch/mips/boot/dts/ralink/mt7621.dtsi b/arch/mips/boot/dts/ralink/mt7621.dtsi
>>> index 35a10258f..dca415fdd 100644
>>> --- a/arch/mips/boot/dts/ralink/mt7621.dtsi
>>> +++ b/arch/mips/boot/dts/ralink/mt7621.dtsi
>>> @@ -123,6 +123,9 @@ serial0: serial@c00 {
>>>               reg-shift = <2>;
>>>               reg-io-width = <4>;
>>>               no-loopback-test;
>>> +
>>> +            pinctrl-names = "default";
>>> +            pinctrl-0 = <&uart1_pins>;
>>>           };
>>>             spi0: spi@b00 {
>>
>> The pins are muxed and can be either UART, or some other function that
>> is supported by the mux: this means that the pinctrl-xxx properties shall
>> *not* go into the SoC dtsi file, but in board dts files instead.
>>
>> Said differently: the usage of the UART pins is board-specific, not SoC-wide.
> 
> Thanks for the explanation. I agree that the pinctrl properties
> would make more sense in a serial node extension in a board's dts,
> but my reason for including them in the SoC's dtsi is due to the
> precedent set with these existing nodes:
> 
>    i2c
>    spi0
>    mmc
>    ethernet
>    pcie
> 
> There is also a default function declared for each of the pin
> groups defined under the pinctrl node. These functions co-incide
> with what is intended for each of those device nodes to function
> correctly, rather than in the alternative GPIO-mode.
> 
> So I thought that sticking with that existing pattern would get
> the least resistance from the community.
> 
> I can imagine how moving the pinctrl node to the board dts, and
> then moving all of the pinctrl properties associated with device
> nodes to their board dts references could be a better separation
> logically.
> 
> What do you recommend?

As a maintainer, this is the logic I follow on the MT7621 device tree
source files regarding the description of pin groups:

- Claim the relevant pin group with the default function (pinctrl-names &
   pinctrl-0) on the node that describes a component of the SoC.

- Keep the node disabled and leave it to the board DTS file to enable it.

I don't disable serial@c00 as we can expect every board to use it. Same
goes for ethernet@...00000. Some boards use the pins on the rgmii2 pin
group as GPIO, so the pinctrl-0 property on ethernet@...00000 is
overwritten on the board DTS file without rgmii2_pins listed.

So I'm fine with this patch as is.

Reviewed-by: Arınç ÜNAL <arinc.unal@...nc9.com>

Arınç