Re: [PATCH 6/6] ARM: dts: aspeed: rainier: Add leds that are on optional PCI cable cards

On Wed, 10 Feb 2021, at 21:46, vishwanatha subbanna wrote:

quoted

On 16-Nov-2020, at 11:43 AM, Joel Stanley [off-list ref] wrote:

On Fri, 13 Nov 2020 at 05:59, Vishwanatha Subbanna
[off-list ref] wrote:

quoted

These are LEDs on the cable cards that plug into PCIE slots.
The LEDs are controlled by PCA9552 I2C expander

Signed-off-by: Vishwanatha Subbanna <redacted>
---
arch/arm/boot/dts/aspeed-bmc-ibm-rainier.dts | 288 +++++++++++++++++++++++++++
1 file changed, 288 insertions(+)

diff --git a/arch/arm/boot/dts/aspeed-bmc-ibm-rainier.dts b/arch/arm/boot/dts/aspeed-bmc-ibm-rainier.dts
index 67c8c40..7de5f76 100644
--- a/arch/arm/boot/dts/aspeed-bmc-ibm-rainier.dts
+++ b/arch/arm/boot/dts/aspeed-bmc-ibm-rainier.dts

@@ -696,6 +696,70 @@
                       gpios = <&pca4 7 GPIO_ACTIVE_LOW>;
               };
       };
+
+       leds-optional-cablecard0 {

Is it necessary to have separate nodes for each of the different GPIO devices?

Would it make sense to combine them, or is it better to be separate?

Andrew, Eddie, Brad: please review this one before I merge it.

I answered this in previous patch set.  If I express ?em all in one 
node that is ?leds {", then if any of the GPIO is not seen because of 
not having the card, then the current leds-gpio driver knocks off all 
the ones on which it successfully acquired the GPIOs also, leaving 
nothing.

I'm struggling to follow this sentence. Can you please explain what you're 
trying to say in a less colloquial way?

I did speak to the maintainer and it looked like the behaviour 
was existing since long time and changing it would break old code.

Break how?

Andrew