Skip to contentSkip to conversation

gpio: vf610: add get_direction() support

PATCH In review [email protected] 2 participants 4 messages Jul 22, 2024 – Jul 29, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gpio: vf610: add get_direction() support

From: Haibo Chen <[email protected]>

For IP which do not contain PDDR, currently use the pinmux API pinctrl_gpio_direction_input() to config the output/input, pinmux currently do not support get_direction(). So here add the GPIO get_direction() support only for the IP which has Port Data Direction Register (PDDR).

---
 drivers/gpio/gpio-vf610.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)
drivers/gpio/gpio-vf610.c +19 -0
index 07e5e6323e86..08ca8377b19c 100644
@@ -151,6 +151,19 @@ static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio,
 	return pinctrl_gpio_direction_output(chip, gpio);
 }
 
+static int vf610_gpio_get_direction(struct gpio_chip *gc, unsigned int gpio)
+{
+	struct vf610_gpio_port *port = gpiochip_get_data(gc);
+	unsigned long mask = BIT(gpio);
+
+	mask &= vf610_gpio_readl(port->gpio_base + GPIO_PDDR);
+
+	if (mask)
+		return GPIO_LINE_DIRECTION_OUT;
+
+	return GPIO_LINE_DIRECTION_IN;
+}
+
 static void vf610_gpio_irq_handler(struct irq_desc *desc)
 {
 	struct vf610_gpio_port *port =
@@ -362,6 +375,12 @@ static int vf610_gpio_probe(struct platform_device *pdev)
 	gc->get = vf610_gpio_get;
 	gc->direction_output = vf610_gpio_direction_output;
 	gc->set = vf610_gpio_set;
+	/*
+	 * only IP has Port Data Direction Register(PDDR) can
+	 * support get direction
+	 */
+	if (port->sdata->have_paddr)
+		gc->get_direction = vf610_gpio_get_direction;
 
 	/* Mask all GPIO interrupts */
 	for (i = 0; i < gc->ngpio; i++)
Signature · 1 line
2.34.1
#1 lore raw

Hi Haibo,

Am 22.07.24 um 08:28 schrieb [email protected]:

Quoted 18 lines

From: Haibo Chen <[email protected]>

For IP which do not contain PDDR, currently use the pinmux API pinctrl_gpio_direction_input() to config the output/input, pinmux currently do not support get_direction(). So here add the GPIO get_direction() support only for the IP which has Port Data Direction Register (PDDR).

---
  drivers/gpio/gpio-vf610.c | 19 +++++++++++++++++++
  1 file changed, 19 insertions(+)
drivers/gpio/gpio-vf610.c +0 -0
index 07e5e6323e86..08ca8377b19c 100644

drivers/gpio/gpio-vf610.c static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio, · line 157

Quoted 7 lines
drivers/gpio/gpio-vf610.c +4 -0
@@ -151,6 +151,19 @@ static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio,
  	return pinctrl_gpio_direction_output(chip, gpio);
  }
 
+static int vf610_gpio_get_direction(struct gpio_chip *gc, unsigned int gpio)
+{
+	struct vf610_gpio_port *port = gpiochip_get_data(gc);
+	unsigned long mask = BIT(gpio);

thanks for sending this patch. I'm fine with this patch, but could we use u32 to make it clear about the range of the mask?

Regards

Quoted 25 lines
drivers/gpio/gpio-vf610.c +15 -0
+
+	mask &= vf610_gpio_readl(port->gpio_base + GPIO_PDDR);
+
+	if (mask)
+		return GPIO_LINE_DIRECTION_OUT;
+
+	return GPIO_LINE_DIRECTION_IN;
+}
+
  static void vf610_gpio_irq_handler(struct irq_desc *desc)
  {
  	struct vf610_gpio_port *port =
@@ -362,6 +375,12 @@ static int vf610_gpio_probe(struct platform_device *pdev)
  	gc->get = vf610_gpio_get;
  	gc->direction_output = vf610_gpio_direction_output;
  	gc->set = vf610_gpio_set;
+	/*
+	 * only IP has Port Data Direction Register(PDDR) can
+	 * support get direction
+	 */
+	if (port->sdata->have_paddr)
+		gc->get_direction = vf610_gpio_get_direction;
 
  	/* Mask all GPIO interrupts */
  	for (i = 0; i < gc->ngpio; i++)
author
#2 lore raw
Quoted 34 lines

-----Original Message-----
From: Stefan Wahren <[email protected]>
Sent: 2024年7月22日 18:32
To: Bough Chen <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] gpio: vf610: add get_direction() support

Hi Haibo,

Am 22.07.24 um 08:28 schrieb [email protected]:

drivers/gpio/gpio-vf610.c static int vf610_gpio_direction_output(struct

Quoted 18 lines

From: Haibo Chen <[email protected]>

For IP which do not contain PDDR, currently use the pinmux API pinctrl_gpio_direction_input() to config the output/input, pinmux currently do not support get_direction(). So here add the GPIO get_direction() support only for the IP which has Port Data Direction Register (PDDR).

---
  drivers/gpio/gpio-vf610.c | 19 +++++++++++++++++++
  1 file changed, 19 insertions(+)
drivers/gpio/gpio-vf610.c +0 -0
index 07e5e6323e86..08ca8377b19c 100644
@@ -151,6 +151,19 @@ static int vf610_gpio_direction_output(struct

gpio_chip *chip, unsigned gpio,

Quoted 3 lines
drivers/gpio/gpio-vf610.c +0 -0
  	return pinctrl_gpio_direction_output(chip, gpio);
  }
 
Quoted 6 lines

drivers/gpio/gpio-vf610.c static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio, · line 157

Quoted 4 lines
drivers/gpio/gpio-vf610.c +4 -0
+static int vf610_gpio_get_direction(struct gpio_chip *gc, unsigned
+int gpio) {
+	struct vf610_gpio_port *port = gpiochip_get_data(gc);
+	unsigned long mask = BIT(gpio);

thanks for sending this patch. I'm fine with this patch, but could we use u32 to make it clear about the range of the mask?

Yes, u32 seems more clear here, but I notice all other place use unsigned long, so I keep the same code style. I go through the history of this driver, seems no specific explanation about the unsigned long. If anyone know the reason, please comment.

Best Regards
Haibo Chen

Quoted 28 lines

Regards

drivers/gpio/gpio-vf610.c static int vf610_gpio_probe(struct platform_device

Quoted 13 lines
drivers/gpio/gpio-vf610.c +9 -0
+
+	mask &= vf610_gpio_readl(port->gpio_base + GPIO_PDDR);
+
+	if (mask)
+		return GPIO_LINE_DIRECTION_OUT;
+
+	return GPIO_LINE_DIRECTION_IN;
+}
+
  static void vf610_gpio_irq_handler(struct irq_desc *desc)
  {
  	struct vf610_gpio_port *port =
@@ -362,6 +375,12 @@ static int vf610_gpio_probe(struct platform_device

*pdev)

Quoted 12 lines
drivers/gpio/gpio-vf610.c +6 -0
  	gc->get = vf610_gpio_get;
  	gc->direction_output = vf610_gpio_direction_output;
  	gc->set = vf610_gpio_set;
+	/*
+	 * only IP has Port Data Direction Register(PDDR) can
+	 * support get direction
+	 */
+	if (port->sdata->have_paddr)
+		gc->get_direction = vf610_gpio_get_direction;
 
  	/* Mask all GPIO interrupts */
  	for (i = 0; i < gc->ngpio; i++)
#3 lore raw

Am 23.07.24 um 04:45 schrieb Bough Chen:

Quoted 34 lines
Quoted 34 lines

-----Original Message-----
From: Stefan Wahren <[email protected]>
Sent: 2024年7月22日 18:32
To: Bough Chen <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH] gpio: vf610: add get_direction() support

Hi Haibo,

Am 22.07.24 um 08:28 schrieb [email protected]:

drivers/gpio/gpio-vf610.c static int vf610_gpio_direction_output(struct

Quoted 18 lines

From: Haibo Chen <[email protected]>

For IP which do not contain PDDR, currently use the pinmux API pinctrl_gpio_direction_input() to config the output/input, pinmux currently do not support get_direction(). So here add the GPIO get_direction() support only for the IP which has Port Data Direction Register (PDDR).

---
   drivers/gpio/gpio-vf610.c | 19 +++++++++++++++++++
   1 file changed, 19 insertions(+)
drivers/gpio/gpio-vf610.c +0 -0
index 07e5e6323e86..08ca8377b19c 100644
@@ -151,6 +151,19 @@ static int vf610_gpio_direction_output(struct

gpio_chip *chip, unsigned gpio,

Quoted 3 lines
drivers/gpio/gpio-vf610.c +0 -0
   	return pinctrl_gpio_direction_output(chip, gpio);
   }
 
Quoted 8 lines
Quoted 6 lines

drivers/gpio/gpio-vf610.c static int vf610_gpio_direction_output(struct gpio_chip *chip, unsigned gpio, · line 157

Quoted 4 lines
drivers/gpio/gpio-vf610.c +4 -0
+static int vf610_gpio_get_direction(struct gpio_chip *gc, unsigned
+int gpio) {
+	struct vf610_gpio_port *port = gpiochip_get_data(gc);
+	unsigned long mask = BIT(gpio);

thanks for sending this patch. I'm fine with this patch, but could we use u32 to make it clear about the range of the mask?

Yes, u32 seems more clear here, but I notice all other place use unsigned long, so I keep the same code style. I go through the history of this driver, seems no specific explanation about the unsigned long.

I understand, but i don't think there is a reason for using unsigned long. So maybe this is a good opportunity to clean this up.

Regards