public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH v1] gpiolib: fix trace on missing gpiochip->get_direction callback
@ 2026-04-09 13:27 Frank Wunderlich
  2026-04-09 14:23 ` Bartosz Golaszewski
  0 siblings, 1 reply; 3+ messages in thread
From: Frank Wunderlich @ 2026-04-09 13:27 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski
  Cc: Frank Wunderlich, linux-gpio, linux-kernel, Daniel Golle

From: Frank Wunderlich <frank-w@public-files.de>

if gpio_chip.get_direction callback is not implemented (e.g. pinctrl-moore) there
is a bunch of traces because of this.

Just remove the WARN_ON to avoid traces and restore previous behaviour but keep the
sanitization active.

Fixes: 471e998c0e31 ("gpiolib: remove redundant callback check")
Fixes: e623c4303ed1 ("gpiolib: sanitize the return value of gpio_chip::get_direction()")
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
---
 drivers/gpio/gpiolib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 86a171e96b0e..302cbd7989f3 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -420,7 +420,7 @@ static int gpiochip_get_direction(struct gpio_chip *gc, unsigned int offset)
 
 	lockdep_assert_held(&gc->gpiodev->srcu);
 
-	if (WARN_ON(!gc->get_direction))
+	if (!gc->get_direction)
 		return -EOPNOTSUPP;
 
 	ret = gc->get_direction(gc, offset);
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH v1] gpiolib: fix trace on missing gpiochip->get_direction callback
  2026-04-09 13:27 [PATCH v1] gpiolib: fix trace on missing gpiochip->get_direction callback Frank Wunderlich
@ 2026-04-09 14:23 ` Bartosz Golaszewski
  2026-04-09 20:42   ` Aw: " Frank Wunderlich
  0 siblings, 1 reply; 3+ messages in thread
From: Bartosz Golaszewski @ 2026-04-09 14:23 UTC (permalink / raw)
  To: Frank Wunderlich
  Cc: Frank Wunderlich, linux-gpio, linux-kernel, Daniel Golle,
	Linus Walleij, Bartosz Golaszewski

On Thu, 9 Apr 2026 15:27:23 +0200, Frank Wunderlich <linux@fw-web.de> said:
> From: Frank Wunderlich <frank-w@public-files.de>
>
> if gpio_chip.get_direction callback is not implemented (e.g. pinctrl-moore) there
> is a bunch of traces because of this.
>
> Just remove the WARN_ON to avoid traces and restore previous behaviour but keep the
> sanitization active.
>
> Fixes: 471e998c0e31 ("gpiolib: remove redundant callback check")
> Fixes: e623c4303ed1 ("gpiolib: sanitize the return value of gpio_chip::get_direction()")
> Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
> ---
>  drivers/gpio/gpiolib.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
> index 86a171e96b0e..302cbd7989f3 100644
> --- a/drivers/gpio/gpiolib.c
> +++ b/drivers/gpio/gpiolib.c
> @@ -420,7 +420,7 @@ static int gpiochip_get_direction(struct gpio_chip *gc, unsigned int offset)
>
>  	lockdep_assert_held(&gc->gpiodev->srcu);
>
> -	if (WARN_ON(!gc->get_direction))
> +	if (!gc->get_direction)
>  		return -EOPNOTSUPP;
>
>  	ret = gc->get_direction(gc, offset);
> --
> 2.43.0
>
>

I prefer GPIO drivers to just implement get_direction(). Looking at the code
it should be pretty straightforward for this driver.

Can you test if the following works for you?

diff --git a/drivers/pinctrl/mediatek/pinctrl-moore.c
b/drivers/pinctrl/mediatek/pinctrl-moore.c
index 70f608347a5f6..071ba849e5322 100644
--- a/drivers/pinctrl/mediatek/pinctrl-moore.c
+++ b/drivers/pinctrl/mediatek/pinctrl-moore.c
@@ -520,6 +520,23 @@ static int mtk_gpio_direction_output(struct
gpio_chip *chip, unsigned int gpio,
 	return pinctrl_gpio_direction_output(chip, gpio);
 }

+static int mtk_gpio_get_direction(struct gpio_chip *chip, unsigned int offset)
+{
+	struct mtk_pinctrl *hw = gpiochip_get_data(chip);
+	const struct mtk_pin_desc *desc;
+	int ret, dir;
+
+	desc = (const struct mtk_pin_desc *)&hw->soc->pins[offset];
+	if (!desc->name)
+		return -ENOTSUPP;
+
+	ret = mtk_hw_get_value(hw, desc, PINCTRL_PIN_REG_DIR, &dir);
+	if (ret)
+		return ret;
+
+	return dir ? GPIO_LINE_DIRECTION_OUT : GPIO_LINE_DIRECTION_IN;
+}
+
 static int mtk_gpio_to_irq(struct gpio_chip *chip, unsigned int offset)
 {
 	struct mtk_pinctrl *hw = gpiochip_get_data(chip);
@@ -566,6 +583,7 @@ static int mtk_build_gpiochip(struct mtk_pinctrl *hw)
 	chip->parent		= hw->dev;
 	chip->request		= gpiochip_generic_request;
 	chip->free		= gpiochip_generic_free;
+	chip->get_direction	= mtk_gpio_get_direction;
 	chip->direction_input	= pinctrl_gpio_direction_input;
 	chip->direction_output	= mtk_gpio_direction_output;
 	chip->get		= mtk_gpio_get;

Thanks,
Bart

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Aw: Re: [PATCH v1] gpiolib: fix trace on missing gpiochip->get_direction callback
  2026-04-09 14:23 ` Bartosz Golaszewski
@ 2026-04-09 20:42   ` Frank Wunderlich
  0 siblings, 0 replies; 3+ messages in thread
From: Frank Wunderlich @ 2026-04-09 20:42 UTC (permalink / raw)
  To: brgl, linux; +Cc: linux-gpio, linux-kernel, daniel, linusw, brgl

Hi

> Gesendet: Donnerstag, 9. April 2026 um 16:23
> Von: "Bartosz Golaszewski" <brgl@kernel.org>
> 
> I prefer GPIO drivers to just implement get_direction(). Looking at the code
> it should be pretty straightforward for this driver.
> 
> Can you test if the following works for you?
> 
> diff --git a/drivers/pinctrl/mediatek/pinctrl-moore.c
> b/drivers/pinctrl/mediatek/pinctrl-moore.c
> index 70f608347a5f6..071ba849e5322 100644
> --- a/drivers/pinctrl/mediatek/pinctrl-moore.c
> +++ b/drivers/pinctrl/mediatek/pinctrl-moore.c
> @@ -520,6 +520,23 @@ static int mtk_gpio_direction_output(struct
> gpio_chip *chip, unsigned int gpio,
>  	return pinctrl_gpio_direction_output(chip, gpio);
>  }
> 
> +static int mtk_gpio_get_direction(struct gpio_chip *chip, unsigned int offset)
> +{
> +	struct mtk_pinctrl *hw = gpiochip_get_data(chip);
> +	const struct mtk_pin_desc *desc;
> +	int ret, dir;
> +
> +	desc = (const struct mtk_pin_desc *)&amp;hw->soc->pins[offset];
> +	if (!desc->name)
> +		return -ENOTSUPP;
> +
> +	ret = mtk_hw_get_value(hw, desc, PINCTRL_PIN_REG_DIR, &amp;dir);
> +	if (ret)
> +		return ret;
> +
> +	return dir ? GPIO_LINE_DIRECTION_OUT : GPIO_LINE_DIRECTION_IN;
> +}
> +
>  static int mtk_gpio_to_irq(struct gpio_chip *chip, unsigned int offset)
>  {
>  	struct mtk_pinctrl *hw = gpiochip_get_data(chip);
> @@ -566,6 +583,7 @@ static int mtk_build_gpiochip(struct mtk_pinctrl *hw)
>  	chip->parent		= hw->dev;
>  	chip->request		= gpiochip_generic_request;
>  	chip->free		= gpiochip_generic_free;
> +	chip->get_direction	= mtk_gpio_get_direction;
>  	chip->direction_input	= pinctrl_gpio_direction_input;
>  	chip->direction_output	= mtk_gpio_direction_output;
>  	chip->get		= mtk_gpio_get;

it looks good on a quick test, just reverted my patch and applied yours and got no trace 

root@bpi-r4:~# cat /sys/kernel/debug/gpio 
gpiochip0: 84 GPIOs, parent: platform/1001f000.pinctrl, pinctrl_moore:
 gpio-0   (                    |tx-disable          ) in  lo 
 gpio-1   (                    |mod-def0            ) in  hi IRQ ACTIVE LOW
 gpio-2   (                    |los                 ) in  hi IRQ 
 gpio-12  (                    |cd                  ) in  lo IRQ ACTIVE LOW
 gpio-13  (                    |reset               ) in  hi IRQ ACTIVE LOW
 gpio-14  (                    |WPS                 ) in  hi IRQ ACTIVE LOW
 gpio-21  (                    |tx-disable          ) out hi 
 gpio-69  (                    |mod-def0            ) in  lo IRQ ACTIVE LOW
 gpio-70  (                    |los                 ) in  hi IRQ 

gpiochip1: 16 GPIOs, parent: i2c/3-0020, 3-0020, can sleep:
 gpio-14  (                    |blue:activity       ) out lo 
 gpio-15  (                    |red:fault           ) out lo 

output seems correct.

regards Frank

> Thanks,
> Bart

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2026-04-09 20:42 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-04-09 13:27 [PATCH v1] gpiolib: fix trace on missing gpiochip->get_direction callback Frank Wunderlich
2026-04-09 14:23 ` Bartosz Golaszewski
2026-04-09 20:42   ` Aw: " Frank Wunderlich

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox