Re: [PATCH 1/1] ARM: dts: stm32: fix AV96 board SAI2B pin muxing on stm32mp15

From: Marek Vasut
Date: Thu Sep 30 2021 - 11:13:16 EST


On 9/30/21 5:05 PM, Alexandre TORGUE wrote:
On 9/30/21 12:26 PM, Marek Vasut wrote:
On 9/30/21 10:47 AM, Alexandre TORGUE wrote:
Hi Marek

On 9/29/21 1:18 PM, Marek Vasut wrote:
On 9/27/21 1:45 PM, Olivier Moysan wrote:
Fix SAI2B pin muxing for AV96 board on STM32MP15.
The label "sai2a-4" is defined twice. Change redundant label to "sai2b-4".

Fixes: dcf185ca8175 ("ARM: dts: stm32: Add alternate pinmux for SAI2 pins on stm32mp15")

Signed-off-by: Olivier Moysan <olivier.moysan@xxxxxxxxxxx>
---
  arch/arm/boot/dts/stm32mp15-pinctrl.dtsi | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/stm32mp15-pinctrl.dtsi b/arch/arm/boot/dts/stm32mp15-pinctrl.dtsi
index 5b60ecbd718f..b9cc9e0dd4fc 100644
--- a/arch/arm/boot/dts/stm32mp15-pinctrl.dtsi
+++ b/arch/arm/boot/dts/stm32mp15-pinctrl.dtsi
@@ -1235,7 +1235,7 @@
          };
      };
-    sai2b_pins_c: sai2a-4 {
+    sai2b_pins_c: sai2b-4 {
          pins1 {
              pinmux = <STM32_PINMUX('F', 11, AF10)>; /* SAI2_SD_B */
              bias-disable;

This mp1 pinmuxing is a total mess, sigh.

What is the issue here ?

The same-old discussion about where to place the pinmux nodes, whether we should have these clusters of pre-defined options in ...pinctrl.dtsi, or whether we should do more nxp-like per-board configuration.

ok it's a bit more precise. Honestly I don't understand why the current topology is an issue here. Maybe pinctrl SAI nodes names are not well chosen or are not enough explicit. Concerning our topology and the NXP ones both exists and both have advantages and drawbacks. For ST boards (DK/EV) we want to keep all configs in the same place.

As I prefer to not re open this topic again and again, feel free to add your pin config in your dts board file, I'll accept it.

Yes, let's not reopen it. I'm sorry if my grumbling came across as too strong.