Skip to contentSkip to conversation

arm64: mediatek: remove broken pmic interrupt property

PATCH In review Arnd Bergmann 3 participants 4 messages Dec 12, 2023 – Dec 15, 2023

Subsystems (2)
Device tree bindings Arm MediaTek
lore
Share & tools
lore

Arnd Bergmann

author
#0 lore raw

PATCH 1/1 arm64: mediatek: remove broken pmic interrupt property

From: Arnd Bergmann <[email protected]>

The pmic is connected to the GIC, which uses four-cell interrupt properties, but its interrupt is specified as two-cell that would only make sense for the GPIO irqchip:

arch/arm64/boot/dts/mediatek/mt8195.dtsi:464.4-27: Warning (interrupts_property): /soc/i2c@11d01000/pmic@34:#interrupt-cells: size is (8), expected multiple of 16

Remove the interrupt for now to shut up the warning. When someone figures out what the correct interrupt and parent are, we can add it back.

Fixes f2b543a191b6 ("arm64: dts: mediatek: add device-tree for Genio 1200 EVK board")
Signed-off-by Arnd Bergmann <[email protected]>
---
 arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts | 2 --
 1 file changed, 2 deletions(-)
arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts +0 -2
index 70b465f7c6a7..a409ef998746 100644
@@ -238,8 +238,6 @@ &i2c6 {
 	mt6360: pmic@34 {
 		compatible = "mediatek,mt6360";
 		reg = <0x34>;
-		interrupts = <128 IRQ_TYPE_EDGE_FALLING>;
-		interrupt-names = "IRQB";
 		interrupt-controller;
 		#interrupt-cells = <1>;
 		pinctrl-0 = <&mt6360_pins>;
Signature · 1 line
2.39.2
maintainer
#1 lore raw

Il 12/12/23 22:47, Arnd Bergmann ha scritto:

Quoted 11 lines

From: Arnd Bergmann <[email protected]>

The pmic is connected to the GIC, which uses four-cell interrupt properties, but its interrupt is specified as two-cell that would only make sense for the GPIO irqchip:

arch/arm64/boot/dts/mediatek/mt8195.dtsi:464.4-27: Warning (interrupts_property): /soc/i2c@11d01000/pmic@34:#interrupt-cells: size is (8), expected multiple of 16

Remove the interrupt for now to shut up the warning. When someone figures out what the correct interrupt and parent are, we can add it back.

Please, can anyone from MediaTek comment on that?

I see a mt6360_pins on PIO:

			pinmux = <PINMUX_GPIO17__FUNC_GPIO17>,
				 <PINMUX_GPIO128__FUNC_GPIO128>;

...and that's GPIO128, which may effectively be the IRQ pin for MT6360.

Still, I'm not sure whether the interrupt is on GIC or PIO, please clarify, otherwise we will have to get this commit upstream.

Thanks,
Angelo

Quoted 19 lines
Fixes f2b543a191b6 ("arm64: dts: mediatek: add device-tree for Genio 1200 EVK board")
Signed-off-by Arnd Bergmann <[email protected]>
---
  arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts | 2 --
  1 file changed, 2 deletions(-)
arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts +0 -2
index 70b465f7c6a7..a409ef998746 100644
@@ -238,8 +238,6 @@ &i2c6 {
  	mt6360: pmic@34 {
  		compatible = "mediatek,mt6360";
  		reg = <0x34>;
-		interrupts = <128 IRQ_TYPE_EDGE_FALLING>;
-		interrupt-names = "IRQB";
  		interrupt-controller;
  		#interrupt-cells = <1>;
  		pinctrl-0 = <&mt6360_pins>;
#2 lore raw
Quoted 7 lines · AngeloGioacchino Del Regno

Il 12/12/23 22:47, Arnd Bergmann ha scritto:

Quoted 6 lines

From: Arnd Bergmann <[email protected]>

The pmic is connected to the GIC, which uses four-cell interrupt properties, but its interrupt is specified as two-cell that would only make sense for the GPIO irqchip:

Quoted 7 lines
Quoted 5 lines

arch/arm64/boot/dts/mediatek/mt8195.dtsi:464.4-27: Warning (interrupts_property): /soc/i2c@11d01000/pmic@34:#interrupt-cells: size is (8), expected multiple of 16

Remove the interrupt for now to shut up the warning. When someone figures out what the correct interrupt and parent are, we can add it back.

Please, can anyone from MediaTek comment on that?

Sorry for late response, I've just stuck in other tasks and didn't notice this

patch is for genio-1200-evk. I've tested 6.7-rc1 but I thought it might be an issue

in origin mt8195.dtsi and not be related to board dts.

Quoted 9 lines

I see a mt6360_pins on PIO:
            pinmux = <PINMUX_GPIO17__FUNC_GPIO17>, <PINMUX_GPIO128__FUNC_GPIO128>;

...and that's GPIO128, which may effectively be the IRQ pin for MT6360.

Still, I'm not sure whether the interrupt is on GIC or PIO, please clarify, otherwise we will have to get this commit upstream.

I think it could be solved by adding

'interrupt-parent = <&pio>;' to mt6360 node.

But currently I have no much time to verify it.

I'll try to find some time to find out the solution tomorrow.

Quoted 9 lines

Thanks,
Angelo

Quoted 6 lines
Fixes f2b543a191b6 ("arm64: dts: mediatek: add device-tree for Genio 1200 EVK board")
Signed-off-by Arnd Bergmann <[email protected]>

---
  arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts | 2 --

  1 file changed, 2 deletions(-)
Quoted 14 lines

arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts &i2c6 { · line 243

Quoted 13 lines
arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts +0 -2
index 70b465f7c6a7..a409ef998746 100644
@@ -238,8 +238,6 @@ &i2c6 {
       mt6360: pmic@34 {
           compatible = "mediatek,mt6360";
           reg = <0x34>;
-        interrupts = <128 IRQ_TYPE_EDGE_FALLING>;
-        interrupt-names = "IRQB";
           interrupt-controller;
           #interrupt-cells = <1>;
           pinctrl-0 = <&mt6360_pins>;

Thanks for pointing out the issue! Let's see if it could be solved by adding 'interrupt-parents'.

Macpaul Lin

#3 lore raw
Quoted 64 lines · Macpaul Lin
Quoted 7 lines · AngeloGioacchino Del Regno

Il 12/12/23 22:47, Arnd Bergmann ha scritto:

Quoted 6 lines

From: Arnd Bergmann <[email protected]>

The pmic is connected to the GIC, which uses four-cell interrupt properties, but its interrupt is specified as two-cell that would only make sense for the GPIO irqchip:

Quoted 7 lines
Quoted 5 lines

arch/arm64/boot/dts/mediatek/mt8195.dtsi:464.4-27: Warning (interrupts_property): /soc/i2c@11d01000/pmic@34:#interrupt-cells: size is (8), expected multiple of 16

Remove the interrupt for now to shut up the warning. When someone figures out what the correct interrupt and parent are, we can add it back.

Please, can anyone from MediaTek comment on that?

Sorry for late response, I've just stuck in other tasks and didn't notice this

patch is for genio-1200-evk. I've tested 6.7-rc1 but I thought it might be an issue

in origin mt8195.dtsi and not be related to board dts.

Quoted 9 lines

I see a mt6360_pins on PIO:
            pinmux = <PINMUX_GPIO17__FUNC_GPIO17>, <PINMUX_GPIO128__FUNC_GPIO128>;

...and that's GPIO128, which may effectively be the IRQ pin for MT6360.

Still, I'm not sure whether the interrupt is on GIC or PIO, please clarify, otherwise we will have to get this commit upstream.

I think it could be solved by adding

'interrupt-parent = <&pio>;' to mt6360 node.

But currently I have no much time to verify it.

I'll try to find some time to find out the solution tomorrow.

Quoted 9 lines

Thanks,
Angelo

Quoted 6 lines
Fixes f2b543a191b6 ("arm64: dts: mediatek: add device-tree for Genio 1200 EVK board")
Signed-off-by Arnd Bergmann <[email protected]>

---
  arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts | 2 --

  1 file changed, 2 deletions(-)
Quoted 14 lines

arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts &i2c6 { · line 243

Quoted 13 lines
arch/arm64/boot/dts/mediatek/mt8395-genio-1200-evk.dts +0 -2
index 70b465f7c6a7..a409ef998746 100644
@@ -238,8 +238,6 @@ &i2c6 {
       mt6360: pmic@34 {
           compatible = "mediatek,mt6360";
           reg = <0x34>;
-        interrupts = <128 IRQ_TYPE_EDGE_FALLING>;
-        interrupt-names = "IRQB";
           interrupt-controller;
           #interrupt-cells = <1>;
           pinctrl-0 = <&mt6360_pins>;
Quoted 3 lines

Thanks for pointing out the issue! Let's see if it could be solved by adding 'interrupt-parents'.

Macpaul Lin

I've send a patch to add 'interrupt-parent' for mt6360 and tested with my local kernel-6.7-rc1.

Link https://lore.kernel.org/linux-devicetree/[email protected]/ open here

Thanks for Angelo , Arnd, and Chen-yu's reminding.

Best regards,

Macpaul Lin