[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <a483fb50-f978-4e48-b38e-6d79632540f1@foss.st.com>
Date: Tue, 29 Oct 2024 16:39:40 +0100
From: Alexandre TORGUE <alexandre.torgue@...s.st.com>
To: Ahmad Fatoum <a.fatoum@...gutronix.de>,
Oleksij Rempel
<o.rempel@...gutronix.de>,
Maxime Coquelin <mcoquelin.stm32@...il.com>,
Rob
Herring <robh+dt@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor
Dooley <conor+dt@...nel.org>
CC: <devicetree@...r.kernel.org>, <netdev@...r.kernel.org>,
<linux-kernel@...r.kernel.org>, <kernel@...gutronix.de>,
<linux-stm32@...md-mailman.stormreply.com>,
<linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH v1] arm: dts: st: stm32mp151a-prtt1l: Fix QSPI
configuration
Hi Ahmad
On 8/7/24 11:38, Ahmad Fatoum wrote:
> Hello Oleksij,
>
> On 06.08.24 14:05, Oleksij Rempel wrote:
>> Rename 'pins1' to 'pins' in the qspi_bk1_pins_a node to correct the
>> subnode name. The previous name caused the configuration to be
>> applied to the wrong subnode, resulting in QSPI not working properly.
>>
>> Signed-off-by: Oleksij Rempel <o.rempel@...gutronix.de>
>> ---
>> arch/arm/boot/dts/st/stm32mp151a-prtt1l.dtsi | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/arch/arm/boot/dts/st/stm32mp151a-prtt1l.dtsi b/arch/arm/boot/dts/st/stm32mp151a-prtt1l.dtsi
>> index 3938d357e198f..4db684478c320 100644
>> --- a/arch/arm/boot/dts/st/stm32mp151a-prtt1l.dtsi
>> +++ b/arch/arm/boot/dts/st/stm32mp151a-prtt1l.dtsi
>> @@ -123,7 +123,7 @@ flash@0 {
>> };
>>
>> &qspi_bk1_pins_a {
>> - pins1 {
>> + pins {
>
> As you have seen such device tree overriding is error prone and would
> be entirely avoidable if specifying full board-specific pinctrl groups
> was allowed for the stm32 platforms instead of override-and-pray.
>
> Anyways, there's better syntax for such overriding now:
>
> &{qspi_blk1_pins_a/pins}
>
> which would cause a compilation error if pins was renamed again.
>
>> bias-pull-up;
>
> There's bias-disable in stm32mp15-pinctrl.dtsi. You may want to add
> a /delete-property/ for that to make sure, it's not up to the driver
> which one has priority.
>
>> drive-push-pull;
>> slew-rate = <1>;
>
> These are already in qspi_bk1_pins_a. If repeating those is ok, why
> not go a step further and just duplicate the pinmux property and stay
> clear of this issue altogether, provided Alex is amenable to changing
> his mind regarding pinctrl groups in board device trees.
I still prefer to have pin configuration defined in pinctrl dtsi file
and I'll continue like this for ST board. The reason is that we try to
reuse as much as possible pins when we create a new board and so it is
easier to maintain if we declare them only one time.
But, I'm not blocked for "other" boards based on STM32 SoC. I mean, if
it is simpler for you and above all if it avoid issues/complexities
then, you can declare some pin groups in your board dts file. In this
case we need to take care of the IO groups label name.
regards
alex
>
>
> Cheers,
> Ahmad
>
Powered by blists - more mailing lists