Skip to contentSkip to conversation

backlight: gpio: add support for multiple GPIOs for backlight control

PATCH In review Sudarshan Shetty 3 participants 12 messages Jan 20 – Feb 2

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

v2 cover backlight: gpio: add support for multiple GPIOs for backlight control

Hi all,

This patch extends the gpio-backlight driver and its Device Tree bindings to support multiple GPIOs for controlling a single backlight device.

Some panels require more than one GPIO to enable or disable the backlight, and previously the driver only supported a single GPIO. With this change:
- The driver now handles an array of GPIOs and updates all of them based on brightness state.
- The Device Tree binding has been updated to allow specifying one or more GPIOs for a gpio-backlight node.

This approach avoids describing multiple backlight devices in DT for a single panel.

Changes in v2:
- Used devm_gpiod_get_array() and struct gpio_descs
- Replaced per-index GPIO handling with descriptor array access
- Moved the bitmap allocation to probe using devm_kcalloc().
- Updated commit messages.

Thanks,
Anusha

Sudarshan Shetty (2):
dt-bindings: backlight: gpio-backlight: allow multiple GPIOs
backlight: gpio: add support for multiple GPIOs for backlight control

 .../leds/backlight/gpio-backlight.yaml        | 24 ++++++-
 drivers/video/backlight/gpio_backlight.c      | 66 +++++++++++++------
 2 files changed, 67 insertions(+), 23 deletions(-)
Signature · 1 line
2.34.1
#1 lore raw

PATCH v2 1/2 dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +22 -2
index 584030b6b0b9..4e4a856cbcd7 100644
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer
 
   default-on:
     description: enable the backlight at boot.
@@ -38,4 +48,14 @@ examples:
         default-on;
     };
 
+  - |
+    #include <dt-bindings/gpio/gpio.h>
+    backlight {
+      compatible = "gpio-backlight";
+      gpios = <&gpio3 4 GPIO_ACTIVE_HIGH>,
+              <&gpio3 5 GPIO_ACTIVE_HIGH>,
+              <&gpio3 6 GPIO_ACTIVE_HIGH>;
+      default-on;
+    };
+
 ...
Signature · 1 line
2.34.1
#10 lore raw

PATCH v2 2/2 backlight: gpio: add support for multiple GPIOs for backlight control

In reply to Sudarshan Shetty

The gpio-backlight driver currently supports only a single GPIO to enable or disable a backlight device. Some panels require multiple enable GPIOs to be asserted together.

Extend the driver to support an array of GPIOs for a single backlight instance. All configured GPIOs are toggled together based on the backlight state.

Existing single-GPIO Device Tree users remain unaffected.

---
 drivers/video/backlight/gpio_backlight.c | 66 ++++++++++++++++--------
 1 file changed, 45 insertions(+), 21 deletions(-)
drivers/video/backlight/gpio_backlight.c +45 -21
index 728a546904b0..11d21de82cf5 100644
@@ -14,17 +14,29 @@
 #include <linux/platform_device.h>
 #include <linux/property.h>
 #include <linux/slab.h>
+#include <linux/bitmap.h>
 
 struct gpio_backlight {
 	struct device *dev;
-	struct gpio_desc *gpiod;
+	struct gpio_descs *gpiods;
+	unsigned long *bitmap;
 };
 
 static int gpio_backlight_update_status(struct backlight_device *bl)
 {
 	struct gpio_backlight *gbl = bl_get_data(bl);
+	unsigned int n = gbl->gpiods->ndescs;
+	int br = backlight_get_brightness(bl);
 
-	gpiod_set_value_cansleep(gbl->gpiod, backlight_get_brightness(bl));
+	if (br)
+		bitmap_fill(gbl->bitmap, n);
+	else
+		bitmap_zero(gbl->bitmap, n);
+
+	gpiod_set_array_value_cansleep(n,
+				       gbl->gpiods->desc,
+				       gbl->gpiods->info,
+				       gbl->bitmap);
 
 	return 0;
 }
@@ -48,26 +60,43 @@ static int gpio_backlight_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct gpio_backlight_platform_data *pdata = dev_get_platdata(dev);
 	struct device_node *of_node = dev->of_node;
-	struct backlight_properties props;
+	struct backlight_properties props = { };
 	struct backlight_device *bl;
 	struct gpio_backlight *gbl;
-	int ret, init_brightness, def_value;
+	bool def_value;
+	enum gpiod_flags flags;
+	unsigned int n;
+	int words;
 
-	gbl = devm_kzalloc(dev, sizeof(*gbl), GFP_KERNEL);
-	if (gbl == NULL)
+	gbl = devm_kcalloc(dev, 1, sizeof(*gbl), GFP_KERNEL);
+	if (!gbl)
 		return -ENOMEM;
 
 	if (pdata)
 		gbl->dev = pdata->dev;
 
 	def_value = device_property_read_bool(dev, "default-on");
+	flags = def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW;
+
+	gbl->gpiods = devm_gpiod_get_array(dev, NULL, flags);
+	if (IS_ERR(gbl->gpiods)) {
+		if (PTR_ERR(gbl->gpiods) == -ENODEV)
+			return dev_err_probe(dev, -EINVAL,
+			"The gpios parameter is missing or invalid\n");
+		return PTR_ERR(gbl->gpiods);
+	}
 
-	gbl->gpiod = devm_gpiod_get(dev, NULL, GPIOD_ASIS);
-	if (IS_ERR(gbl->gpiod))
-		return dev_err_probe(dev, PTR_ERR(gbl->gpiod),
-				     "The gpios parameter is missing or invalid\n");
+	n = gbl->gpiods->ndescs;
+	if (!n)
+		return dev_err_probe(dev, -EINVAL,
+			"No GPIOs provided\n");
+
+	words = BITS_TO_LONGS(n);
+	gbl->bitmap = devm_kcalloc(dev, words, sizeof(unsigned long),
+				   GFP_KERNEL);
+	if (!gbl->bitmap)
+		return -ENOMEM;
 
-	memset(&props, 0, sizeof(props));
 	props.type = BACKLIGHT_RAW;
 	props.max_brightness = 1;
 	bl = devm_backlight_device_register(dev, dev_name(dev), dev, gbl,
@@ -81,21 +110,16 @@ static int gpio_backlight_probe(struct platform_device *pdev)
 	if (!of_node || !of_node->phandle)
 		/* Not booted with device tree or no phandle link to the node */
 		bl->props.power = def_value ? BACKLIGHT_POWER_ON
-					    : BACKLIGHT_POWER_OFF;
-	else if (gpiod_get_value_cansleep(gbl->gpiod) == 0)
+						    : BACKLIGHT_POWER_OFF;
+	else if (gpiod_get_value_cansleep(gbl->gpiods->desc[0]) == 0)
 		bl->props.power = BACKLIGHT_POWER_OFF;
 	else
 		bl->props.power = BACKLIGHT_POWER_ON;
 
-	bl->props.brightness = 1;
-
-	init_brightness = backlight_get_brightness(bl);
-	ret = gpiod_direction_output(gbl->gpiod, init_brightness);
-	if (ret) {
-		dev_err(dev, "failed to set initial brightness\n");
-		return ret;
-	}
+	bl->props.brightness = def_value ? 1 : 0;
 
+	gpio_backlight_update_status(bl);
+
 	platform_set_drvdata(pdev, bl);
 	return 0;
 }
Signature · 1 line
2.34.1
#2 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

In reply to Sudarshan Shetty
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

Quoted 11 lines
---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +0 -0
index 584030b6b0b9..4e4a856cbcd7 100644

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 23

Quoted 10 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +5 -2
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array

There is no such syntax in the bindings, from where did you get it? Type is already defined.

items:
minItems: 1
maxItems: 3

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 30

Quoted 7 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +7 -0
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer

All this is some odd stuff - just to be clear, don't send us LLM output. I don't want to waste my time to review microslop.

Was it done with help of Microslop?

Best regards,
Krzysztof

#3 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 38 lines
Quoted 11 lines
---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +0 -0
index 584030b6b0b9..4e4a856cbcd7 100644

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 23

Quoted 10 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +5 -2
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array

There is no such syntax in the bindings, from where did you get it? Type is already defined.

items:
minItems: 1
maxItems: 3

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 30

Quoted 7 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +7 -0
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer
Quoted 5 lines

All this is some odd stuff - just to be clear, don't send us LLM output. I don't want to waste my time to review microslop.

Was it done with help of Microslop?

I understand now that the schema changes I proposed were not correct, and I will address this in the next patch series. My intention was to check whether the gpio-backlight binding could support more than one enable-type GPIO.
Could you please advise what would be an appropriate maximum number of GPIOs for gpio-backlight in such a scenario? For example, would allowing 2 GPIOs be acceptable, or should this case be handled in a different way?

Quoted 2 lines

Best regards,
Krzysztof

#4 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

Quoted 65 lines · On 23-01-2026 16:41, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 38 lines
Quoted 11 lines
---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +0 -0
index 584030b6b0b9..4e4a856cbcd7 100644

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 23

Quoted 10 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +5 -2
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array

There is no such syntax in the bindings, from where did you get it? Type is already defined.

items:
minItems: 1
maxItems: 3

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 30

Quoted 7 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +7 -0
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer
Quoted 5 lines

All this is some odd stuff - just to be clear, don't send us LLM output. I don't want to waste my time to review microslop.

Was it done with help of Microslop?

Quoted 8 lines

I understand now that the schema changes I proposed were not correct, and I will address this in the next patch series. My intention was to check whether the gpio-backlight binding could support more than one enable-type GPIO.
Could you please advise what would be an appropriate maximum number of GPIOs for gpio-backlight in such a scenario? For example, would allowing 2 GPIOs be acceptable, or should this case be handled in a different way?

In line with Daniel’s suggestion, I am planning to adopt a fixed upper limit for the number of backlight GPIOs. The current hardware only requires two GPIOs, so the maxItems can be set to 2.

If future platforms or customers require support for a higher number of GPIOs, this limit can be increased and the driver can be updated accordingly.

Kindly advise if this solution aligns with your expectations, or if you prefer an alternative maximum value.

Quoted 3 lines
Quoted 2 lines

Best regards,
Krzysztof

#6 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

In reply to [email protected]
Quoted 13 lines · On 23/01/2026 12:11, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

Quoted 7 lines

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

That's not an enable gpio, but PWM.

You write bindings for this device, not for something else - like your board.

Quoted 39 lines
Quoted 39 lines
Quoted 12 lines
---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +0 -0
index 584030b6b0b9..4e4a856cbcd7 100644

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 23

Quoted 10 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +5 -2
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array

There is no such syntax in the bindings, from where did you get it? Type is already defined.

items:
minItems: 1
maxItems: 3

Quoted 7 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +7 -0
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer
Quoted 7 lines
Quoted 5 lines

All this is some odd stuff - just to be clear, don't send us LLM output. I don't want to waste my time to review microslop.

Was it done with help of Microslop?

I understand now that the schema changes I proposed were not correct,

How such code could be even created... Just in case, do you understand that Microslop and LLM is waste of our time?

Quoted 6 lines

and I will address this in the next patch series. My intention was to check whether the gpio-backlight binding could support more than one enable-type GPIO.
Could you please advise what would be an appropriate maximum number of GPIOs for gpio-backlight in such a scenario? For example, would allowing 2 GPIOs be acceptable, or should this case be handled in a different way?

We have plenty of examples for this, but anyway you won't need it because this is not an enable GPIO.

Best regards,
Krzysztof

#5 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

In reply to [email protected]
Quoted 81 lines · On 27/01/2026 13:46, [email protected] wrote:
Quoted 65 lines · On 23-01-2026 16:41, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 38 lines
Quoted 11 lines
---
 .../leds/backlight/gpio-backlight.yaml        | 24 +++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +0 -0
index 584030b6b0b9..4e4a856cbcd7 100644

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 23

Quoted 10 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +5 -2
@@ -16,8 +16,18 @@ properties:
     const: gpio-backlight
 
   gpios:
-    description: The gpio that is used for enabling/disabling the backlight.
-    maxItems: 1
+    description: |
+      The gpio that is used for enabling/disabling the backlight.
+      Multiple GPIOs can be specified for panels that require several
+      enable signals. All GPIOs are controlled together.
+    type: array

There is no such syntax in the bindings, from where did you get it? Type is already defined.

items:
minItems: 1
maxItems: 3

Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml properties: · line 30

Quoted 7 lines
Documentation/devicetree/bindings/leds/backlight/gpio-backlight.yaml +7 -0
+    minItems: 1
+    items:
+      type: array
+      minItems: 3
+      maxItems: 3
+      items:
+        type: integer
Quoted 5 lines

All this is some odd stuff - just to be clear, don't send us LLM output. I don't want to waste my time to review microslop.

Was it done with help of Microslop?

Quoted 8 lines

I understand now that the schema changes I proposed were not correct, and I will address this in the next patch series. My intention was to check whether the gpio-backlight binding could support more than one enable-type GPIO.
Could you please advise what would be an appropriate maximum number of GPIOs for gpio-backlight in such a scenario? For example, would allowing 2 GPIOs be acceptable, or should this case be handled in a different way?

In line with Daniel’s suggestion, I am planning to adopt a fixed upper limit for the number of backlight GPIOs. The current hardware only requires two GPIOs, so the maxItems can be set to 2.

Quoted 6 lines

If future platforms or customers require support for a higher number of GPIOs, this limit can be increased and the driver can be updated accordingly.

Kindly advise if this solution aligns with your expectations, or if you prefer an alternative maximum value.

You have entire commit msg to explain the hardware and explain WHY you are doing this. In a concise and readable way. I will not be going through 2 different email threads with 20 messages to figure that out.

Best regards,
Krzysztof

#11 lore raw
In reply to Sudarshan Shetty
Quoted 52 lines · Sudarshan Shetty

The gpio-backlight driver currently supports only a single GPIO to enable or disable a backlight device. Some panels require multiple enable GPIOs to be asserted together.

Extend the driver to support an array of GPIOs for a single backlight instance. All configured GPIOs are toggled together based on the backlight state.

Existing single-GPIO Device Tree users remain unaffected.

---
 drivers/video/backlight/gpio_backlight.c | 66 ++++++++++++++++--------
 1 file changed, 45 insertions(+), 21 deletions(-)
drivers/video/backlight/gpio_backlight.c +14 -2
index 728a546904b0..11d21de82cf5 100644
@@ -14,17 +14,29 @@
 #include <linux/platform_device.h>
 #include <linux/property.h>
 #include <linux/slab.h>
+#include <linux/bitmap.h>
 
 struct gpio_backlight {
 	struct device *dev;
-	struct gpio_desc *gpiod;
+	struct gpio_descs *gpiods;
+	unsigned long *bitmap;
 };
 
 static int gpio_backlight_update_status(struct backlight_device *bl)
 {
 	struct gpio_backlight *gbl = bl_get_data(bl);
+	unsigned int n = gbl->gpiods->ndescs;
+	int br = backlight_get_brightness(bl);
 
-	gpiod_set_value_cansleep(gbl->gpiod, backlight_get_brightness(bl));
+	if (br)
+		bitmap_fill(gbl->bitmap, n);
+	else
+		bitmap_zero(gbl->bitmap, n);
+
+	gpiod_set_array_value_cansleep(n,
+				       gbl->gpiods->desc,
+				       gbl->gpiods->info,
+				       gbl->bitmap);
 
 	return 0;
 }

drivers/video/backlight/gpio_backlight.c static int gpio_backlight_probe(struct platform_device *pdev) · line 63

Quoted 5 lines
drivers/video/backlight/gpio_backlight.c +1 -1
@@ -48,26 +60,43 @@ static int gpio_backlight_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct gpio_backlight_platform_data *pdata = dev_get_platdata(dev);
 	struct device_node *of_node = dev->of_node;
-	struct backlight_properties props;
+	struct backlight_properties props = { };

This change is unrelated to the patch description. Do not "hide" changes like this. It you want to replace the memset() it's better to send a separate patch.

drivers/video/backlight/gpio_backlight.c static int gpio_backlight_probe(struct platform_device *pdev) · line 72

Quoted 12 lines
drivers/video/backlight/gpio_backlight.c +6 -3
 	struct backlight_device *bl;
 	struct gpio_backlight *gbl;
-	int ret, init_brightness, def_value;
+	bool def_value;
+	enum gpiod_flags flags;
+	unsigned int n;
+	int words;
 
-	gbl = devm_kzalloc(dev, sizeof(*gbl), GFP_KERNEL);
-	if (gbl == NULL)
+	gbl = devm_kcalloc(dev, 1, sizeof(*gbl), GFP_KERNEL);
+	if (!gbl)

Again, this change is unrelated to the patch description. Do not include changes that are not described in the patch description.

drivers/video/backlight/gpio_backlight.c static int gpio_backlight_probe(struct platform_device *pdev) · line 81

Quoted 9 lines
drivers/video/backlight/gpio_backlight.c +3 -0
 		return -ENOMEM;
 
 	if (pdata)
 		gbl->dev = pdata->dev;
 
 	def_value = device_property_read_bool(dev, "default-on");
+	flags = def_value ? GPIOD_OUT_HIGH : GPIOD_OUT_LOW;
+
+	gbl->gpiods = devm_gpiod_get_array(dev, NULL, flags);

How is it safe to transition from GPIOD_ASIS to GPIOD_OUT_LOW here?

Forcing the backlight off if the default-on property is not present will prevent the backlight state being properly inherited from the bootloader.

Quoted 27 lines
drivers/video/backlight/gpio_backlight.c +16 -5
+	if (IS_ERR(gbl->gpiods)) {
+		if (PTR_ERR(gbl->gpiods) == -ENODEV)
+			return dev_err_probe(dev, -EINVAL,
+			"The gpios parameter is missing or invalid\n");
+		return PTR_ERR(gbl->gpiods);
+	}
 
-	gbl->gpiod = devm_gpiod_get(dev, NULL, GPIOD_ASIS);
-	if (IS_ERR(gbl->gpiod))
-		return dev_err_probe(dev, PTR_ERR(gbl->gpiod),
-				     "The gpios parameter is missing or invalid\n");
+	n = gbl->gpiods->ndescs;
+	if (!n)
+		return dev_err_probe(dev, -EINVAL,
+			"No GPIOs provided\n");
+
+	words = BITS_TO_LONGS(n);
+	gbl->bitmap = devm_kcalloc(dev, words, sizeof(unsigned long),
+				   GFP_KERNEL);
+	if (!gbl->bitmap)
+		return -ENOMEM;
 
-	memset(&props, 0, sizeof(props));
 	props.type = BACKLIGHT_RAW;
 	props.max_brightness = 1;
 	bl = devm_backlight_device_register(dev, dev_name(dev), dev, gbl,

drivers/video/backlight/gpio_backlight.c static int gpio_backlight_probe(struct platform_device *pdev) · line 114

Quoted 7 lines
drivers/video/backlight/gpio_backlight.c +2 -2
@@ -81,21 +110,16 @@ static int gpio_backlight_probe(struct platform_device *pdev)
 	if (!of_node || !of_node->phandle)
 		/* Not booted with device tree or no phandle link to the node */
 		bl->props.power = def_value ? BACKLIGHT_POWER_ON
-					    : BACKLIGHT_POWER_OFF;
-	else if (gpiod_get_value_cansleep(gbl->gpiod) == 0)
+						    : BACKLIGHT_POWER_OFF;
+	else if (gpiod_get_value_cansleep(gbl->gpiods->desc[0]) == 0)

This logic is broken. This code path needs to be taken is *any* GPIO is low (and, as mentioned, the initial GPIO state must be GPIOD_ASIS).

Quoted 9 lines
drivers/video/backlight/gpio_backlight.c +0 -5
 		bl->props.power = BACKLIGHT_POWER_OFF;
 	else
 		bl->props.power = BACKLIGHT_POWER_ON;
 
-	bl->props.brightness = 1;
-
-	init_brightness = backlight_get_brightness(bl);
-	ret = gpiod_direction_output(gbl->gpiod, init_brightness);
-	if (ret) {

drivers/video/backlight/gpio_backlight.c static int gpio_backlight_probe(struct platform_device *pdev) · line 125

Quoted 10 lines
drivers/video/backlight/gpio_backlight.c +3 -3
-		dev_err(dev, "failed to set initial brightness\n");
-		return ret;
-	}
+	bl->props.brightness = def_value ? 1 : 0;
 
+	gpio_backlight_update_status(bl);
+
 	platform_set_drvdata(pdev, bl);
 	return 0;
 }

Daniel.

#7 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

In reply to Krzysztof Kozlowski
Quoted 22 lines · Krzysztof Kozlowski
Quoted 13 lines · On 23/01/2026 12:11, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

Quoted 7 lines

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 4 lines

That's not an enable gpio, but PWM.

You write bindings for this device, not for something else - like your board.

Sudarshan: I believe at one point the intent was to model this hardware as a pwm-backlight (using enables GPIOs to drive the enable pin) attached to a pwm-gpio (to drive the PWM pin). Did this approach work?

Daniel.

#8 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

Quoted 23 lines · Daniel Thompson
Quoted 22 lines · Krzysztof Kozlowski
Quoted 20 lines · On 23/01/2026 12:11, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 9 lines
Quoted 4 lines

That's not an enable gpio, but PWM.

You write bindings for this device, not for something else - like your board.

Sudarshan: I believe at one point the intent was to model this hardware as a pwm-backlight (using enables GPIOs to drive the enable pin) attached to a pwm-gpio (to drive the PWM pin). Did this approach work?

Yes, the original plan was to model this using pwm‑gpio, and that setup worked. But on the SOC there’s no actual PWM controller available for this path— the LED_PWM line is just tied to a GPIO that’s driven high (effectively a fixed 100% duty cycle). Because of that, describing it as a PWM in DT was flagged as incorrect.

As pointed out during the SoC DTS review, the correct path forward is to extend gpio‑backlight to handle multiple GPIOs, rather than representing them as multiple separate backlight devices.

Quoted 2 lines

Daniel.

#9 lore raw

Re: [PATCH v2 1/2] dt-bindings: backlight: gpio-backlight: allow multiple GPIOs

Quoted 40 lines · On Thu, Jan 29, 2026 at 11:11:34AM +0530, [email protected] wrote:
Quoted 23 lines · Daniel Thompson
Quoted 22 lines · Krzysztof Kozlowski
Quoted 20 lines · On 23/01/2026 12:11, [email protected] wrote:
Quoted 9 lines · Krzysztof Kozlowski
Quoted 2 lines · Sudarshan Shetty

Update the gpio-backlight binding to support configurations that require more than one GPIO for enabling/disabling the backlight.

Why? Which devices need it? How a backlight would have three enable GPIOs? I really do not believe, so you need to write proper hardware justification.

To clarify our hardware setup:
the panel requires one GPIO for the backlight enable signal, and it also has a PWM input. Since the QCS615 does not provide a PWM controller for this use case, the PWM input is connected to a GPIO that is driven high to provide a constant 100% duty cycle, as explained in the link below.
https://lore.kernel.org/all/[email protected]/T/#m93ca4e5c7bf055715ed13316d91f0cd544244cf5

Quoted 9 lines
Quoted 4 lines

That's not an enable gpio, but PWM.

You write bindings for this device, not for something else - like your board.

Sudarshan: I believe at one point the intent was to model this hardware as a pwm-backlight (using enables GPIOs to drive the enable pin) attached to a pwm-gpio (to drive the PWM pin). Did this approach work?

Yes, the original plan was to model this using pwm‑gpio, and that setup worked. But on the SOC there’s no actual PWM controller available for this path— the LED_PWM line is just tied to a GPIO that’s driven high (effectively a fixed 100% duty cycle). Because of that, describing it as a PWM in DT was flagged as incorrect.

Quoted 3 lines

As pointed out during the SoC DTS review, the correct path forward is to extend gpio‑backlight to handle multiple GPIOs, rather than representing them as multiple separate backlight devices.

That not quite what I got from the link above. There is a suggestion to use gpio-backlight, but the reason it was flagged is because pwm-gpio was unused... it was not referenced by a pwm-backlight.

Having said that I suspect it is better to model this backlight controller on this board as a gpio-backlight because from a backlight controller point of this that is physically what the controller is composed of (assuming there is not sufficient capacitance on the signal for a software PWM to work at anything other than 0% and 100%). Even if those GPIO signals are connected to the panel's PWM input I'm not sure that's relevant because none of the backlight controller bindings model the panel anyway.

Whatever route you select, you do need to make it clear in the patch description *why* it is correct to model the system as a gpio-backlight. Deferring to (potentially ambiguous) review comments is not sufficient to explain why changing the gpio-backlight bindings are an improvement.

Daniel.