Skip to contentSkip to conversation

gpiolib: acpi: Program debounce when finding GPIO

PATCH In review Mario Limonciello (AMD) 3 participants 5 messages Aug 18, 2025 – Aug 20, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gpiolib: acpi: Program debounce when finding GPIO

When soc-button-array looks up the GPIO to use it calls acpi_find_gpio() which will parse _CRS.

acpi_find_gpio.cold (drivers/gpio/gpiolib-acpi-core.c:953) gpiod_find_and_request (drivers/gpio/gpiolib.c:4598 drivers/gpio/gpiolib.c:4625) gpiod_get_index (drivers/gpio/gpiolib.c:4877)

The GPIO is setup basically, but the debounce information is discarded. The platform will assert what debounce should be in _CRS, so program it at the time it's available.

---
v4:
* Just add a direct call instead
* drop tag
* update commit message

---
 drivers/gpio/gpiolib-acpi-core.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpio/gpiolib-acpi-core.c +6 -0
index 12b24a717e43f..6388e8e363dee 100644
@@ -944,6 +944,7 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
 	bool can_fallback = acpi_can_fallback_to_crs(adev, con_id);
 	struct acpi_gpio_info info;
 	struct gpio_desc *desc;
+	int ret;
 
 	desc = __acpi_find_gpio(fwnode, con_id, idx, can_fallback, &info);
 	if (IS_ERR(desc))
@@ -957,6 +958,11 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
 
 	acpi_gpio_update_gpiod_flags(dflags, &info);
 	acpi_gpio_update_gpiod_lookup_flags(lookupflags, &info);
+	/* ACPI uses hundredths of milliseconds units */
+	ret = gpio_set_debounce_timeout(desc, info.debounce * 10);
+	if (ret)
+		return ERR_PTR(ret);
+
 	return desc;
 }
 
Signature · 1 line
2.43.0
#1 lore raw

Hi,

Quoted 10 lines · Mario Limonciello (AMD)

When soc-button-array looks up the GPIO to use it calls acpi_find_gpio() which will parse _CRS.

acpi_find_gpio.cold (drivers/gpio/gpiolib-acpi-core.c:953) gpiod_find_and_request (drivers/gpio/gpiolib.c:4598 drivers/gpio/gpiolib.c:4625) gpiod_get_index (drivers/gpio/gpiolib.c:4877)

The GPIO is setup basically, but the debounce information is discarded. The platform will assert what debounce should be in _CRS, so program it at the time it's available.

The spec says for GpioInt():

DebounceTimeout is an optional argument specifying the debounce wait time, in hundredths of milliseconds. The bit field name _DBT is automatically created to refer to this portion of the resource descriptor.

I was not sure but wanted to check that if it is left out, does ACPICA fill it with 0? If yes (I would expect so) then this is fine.

Quoted 3 lines

---
v4:

You missed "v4" in the $subject.

Quoted 33 lines

* Just add a direct call instead
* drop tag
* update commit message

---
 drivers/gpio/gpiolib-acpi-core.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpio/gpiolib-acpi-core.c +6 -0
index 12b24a717e43f..6388e8e363dee 100644
@@ -944,6 +944,7 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
 	bool can_fallback = acpi_can_fallback_to_crs(adev, con_id);
 	struct acpi_gpio_info info;
 	struct gpio_desc *desc;
+	int ret;
 
 	desc = __acpi_find_gpio(fwnode, con_id, idx, can_fallback, &info);
 	if (IS_ERR(desc))
@@ -957,6 +958,11 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
 
 	acpi_gpio_update_gpiod_flags(dflags, &info);
 	acpi_gpio_update_gpiod_lookup_flags(lookupflags, &info);
+	/* ACPI uses hundredths of milliseconds units */
+	ret = gpio_set_debounce_timeout(desc, info.debounce * 10);
+	if (ret)
+		return ERR_PTR(ret);
+
 	return desc;
 }
 
Signature · 1 line
2.43.0
#2 lore raw
Quoted 16 lines · Mika Westerberg

Hi,

Quoted 10 lines · Mario Limonciello (AMD)

When soc-button-array looks up the GPIO to use it calls acpi_find_gpio() which will parse _CRS.

acpi_find_gpio.cold (drivers/gpio/gpiolib-acpi-core.c:953) gpiod_find_and_request (drivers/gpio/gpiolib.c:4598 drivers/gpio/gpiolib.c:4625) gpiod_get_index (drivers/gpio/gpiolib.c:4877)

The GPIO is setup basically, but the debounce information is discarded. The platform will assert what debounce should be in _CRS, so program it at the time it's available.

The spec says for GpioInt():

Quoted 7 lines

DebounceTimeout is an optional argument specifying the debounce wait time, in hundredths of milliseconds. The bit field name _DBT is automatically created to refer to this portion of the resource descriptor.

I was not sure but wanted to check that if it is left out, does ACPICA fill it with 0? If yes (I would expect so) then this is fine.

Yeah AFAICT you're right. The ACPI resource is zero'ed out, so if the field was empty it should default to zero.

Quoted 6 lines
Quoted 4 lines

---
v4:

You missed "v4" in the $subject.

Whoops, thanks. If there ends up being a reason to spin I'll send the next one as a v5.

Quoted 34 lines
Quoted 34 lines

* Just add a direct call instead
* drop tag
* update commit message

---
  drivers/gpio/gpiolib-acpi-core.c | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/gpio/gpiolib-acpi-core.c +6 -0
index 12b24a717e43f..6388e8e363dee 100644
@@ -944,6 +944,7 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
  	bool can_fallback = acpi_can_fallback_to_crs(adev, con_id);
  	struct acpi_gpio_info info;
  	struct gpio_desc *desc;
+	int ret;
 
  	desc = __acpi_find_gpio(fwnode, con_id, idx, can_fallback, &info);
  	if (IS_ERR(desc))
@@ -957,6 +958,11 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
 
  	acpi_gpio_update_gpiod_flags(dflags, &info);
  	acpi_gpio_update_gpiod_lookup_flags(lookupflags, &info);
+	/* ACPI uses hundredths of milliseconds units */
+	ret = gpio_set_debounce_timeout(desc, info.debounce * 10);
+	if (ret)
+		return ERR_PTR(ret);
+
  	return desc;
  }
 
Signature · 1 line
2.43.0
Reviewed-by
#3 lore raw
Quoted 22 lines · Mario Limonciello
Quoted 21 lines · Mika Westerberg

Hi,

Quoted 10 lines · Mario Limonciello (AMD)

When soc-button-array looks up the GPIO to use it calls acpi_find_gpio() which will parse _CRS.

acpi_find_gpio.cold (drivers/gpio/gpiolib-acpi-core.c:953) gpiod_find_and_request (drivers/gpio/gpiolib.c:4598 drivers/gpio/gpiolib.c:4625) gpiod_get_index (drivers/gpio/gpiolib.c:4877)

The GPIO is setup basically, but the debounce information is discarded. The platform will assert what debounce should be in _CRS, so program it at the time it's available.

The spec says for GpioInt():

DebounceTimeout is an optional argument specifying the debounce wait time, in hundredths of milliseconds. The bit field name _DBT is automatically created to refer to this portion of the resource descriptor.

Quoted 5 lines
Quoted 2 lines

I was not sure but wanted to check that if it is left out, does ACPICA fill it with 0? If yes (I would expect so) then this is fine.

Yeah AFAICT you're right. The ACPI resource is zero'ed out, so if the field was empty it should default to zero.

Okay good.

Then from my perspective,

Quoted 41 lines
Quoted 6 lines
Quoted 4 lines

---
v4:

You missed "v4" in the $subject.

Whoops, thanks. If there ends up being a reason to spin I'll send the next one as a v5.

Quoted 31 lines
Quoted 31 lines

* Just add a direct call instead
* drop tag
* update commit message

---
  drivers/gpio/gpiolib-acpi-core.c | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/gpio/gpiolib-acpi-core.c +6 -0
index 12b24a717e43f..6388e8e363dee 100644
@@ -944,6 +944,7 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
  	bool can_fallback = acpi_can_fallback_to_crs(adev, con_id);
  	struct acpi_gpio_info info;
  	struct gpio_desc *desc;
+	int ret;
  	desc = __acpi_find_gpio(fwnode, con_id, idx, can_fallback, &info);
  	if (IS_ERR(desc))
@@ -957,6 +958,11 @@ struct gpio_desc *acpi_find_gpio(struct fwnode_handle *fwnode,
  	acpi_gpio_update_gpiod_flags(dflags, &info);
  	acpi_gpio_update_gpiod_lookup_flags(lookupflags, &info);
+	/* ACPI uses hundredths of milliseconds units */
+	ret = gpio_set_debounce_timeout(desc, info.debounce * 10);
+	if (ret)
+		return ERR_PTR(ret);
+
  	return desc;
  }
Signature · 1 line
2.43.0
maintainer
#4 lore raw
Quoted 3 lines · Mika Westerberg
Quoted 2 lines · Mario Limonciello
Quoted 1 line · Mika Westerberg

On Sun, Aug 17, 2025 at 08:52:07PM -0500, Mario Limonciello (AMD) wrote:

...

Quoted 11 lines
Quoted 5 lines
Quoted 2 lines

I was not sure but wanted to check that if it is left out, does ACPICA fill it with 0? If yes (I would expect so) then this is fine.

Yeah AFAICT you're right. The ACPI resource is zero'ed out, so if the field was empty it should default to zero.

Okay good.

Then from my perspective,

Seems I came from vacation just in time :-) Please, send v5 with all nits and tags and I will apply it.

Signature · 2 lines
With Best Regards,
Andy Shevchenko