[PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Let's start working towards removing gpiod_to_chip() from the kernel.
Start with pinctrl. The first patch should go through the GPIO tree
and become available through an immutable tag so that we can get more
conversions in this cycle elsewhere.

Bartosz Golaszewski (3):
  gpiolib: provide gpio_devie_get_label()
  pinctrl: stop using gpiod_to_chip()
  pinctrl: don't include GPIOLIB private header

 drivers/gpio/gpiolib.c      | 14 ++++++++++++++
 drivers/pinctrl/core.c      | 18 ++++++++----------
 include/linux/gpio/driver.h |  1 +
 3 files changed, 23 insertions(+), 10 deletions(-)

-- 
2.40.1

[PATCH 1/3] gpiolib: provide gpio_devie_get_label()

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Provide a getter for the GPIO device label string so that users don't
have to dereference struct gpio_chip directly.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
---
 drivers/gpio/gpiolib.c      | 14 ++++++++++++++
 include/linux/gpio/driver.h |  1 +
 2 files changed, 15 insertions(+)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 95d2a7b2ea3e..a5faaea6915d 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -254,6 +254,20 @@ int gpio_device_get_base(struct gpio_device *gdev)
 }
 EXPORT_SYMBOL_GPL(gpio_device_get_base);
 
+/**
+ * gpio_device_get_label() - Get the label of this GPIO device
+ * @gdev: GPIO device
+ *
+ * Returns:
+ * Pointer to the string containing the GPIO device label. The string's
+ * lifetime is tied to that of the underlying GPIO device.
+ */
+const char *gpio_device_get_label(struct gpio_device *gdev)
+{
+	return gdev->label;
+}
+EXPORT_SYMBOL(gpio_device_get_label);
+
 /**
  * gpio_device_get_chip() - Get the gpio_chip implementation of this GPIO device
  * @gdev: GPIO device
diff --git a/include/linux/gpio/driver.h b/include/linux/gpio/driver.h
index 0aed62f0c633..100c329dc986 100644
--- a/include/linux/gpio/driver.h
+++ b/include/linux/gpio/driver.h
@@ -786,6 +786,7 @@ struct gpio_device *gpiod_to_gpio_device(struct gpio_desc *desc);
 
 /* struct gpio_device getters */
 int gpio_device_get_base(struct gpio_device *gdev);
+const char *gpio_device_get_label(struct gpio_device *gdev);
 
 #else /* CONFIG_GPIOLIB */
 
-- 
2.40.1

[PATCH 2/3] pinctrl: stop using gpiod_to_chip()

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Don't dereference struct gpio_chip directly, use dedicated gpio_device
getters instead.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
---
 drivers/pinctrl/core.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/pinctrl/core.c b/drivers/pinctrl/core.c
index 1fa89be29b8f..0b4d07aea387 100644
--- a/drivers/pinctrl/core.c
+++ b/drivers/pinctrl/core.c
@@ -13,6 +13,7 @@
 #define pr_fmt(fmt) "pinctrl core: " fmt
 
 #include <linux/array_size.h>
+#include <linux/cleanup.h>
 #include <linux/debugfs.h>
 #include <linux/device.h>
 #include <linux/err.h>
@@ -1649,8 +1650,8 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
 	const struct pinctrl_ops *ops = pctldev->desc->pctlops;
 	unsigned i, pin;
 #ifdef CONFIG_GPIOLIB
+	struct gpio_device *gdev __free(gpio_device_put) = NULL;
 	struct pinctrl_gpio_range *range;
-	struct gpio_chip *chip;
 	int gpio_num;
 #endif
 
@@ -1685,11 +1686,11 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
 			 * we need to get rid of the range->base eventually and
 			 * get the descriptor directly from the gpio_chip.
 			 */
-			chip = gpiod_to_chip(gpio_to_desc(gpio_num));
-		else
-			chip = NULL;
-		if (chip)
-			seq_printf(s, "%u:%s ", gpio_num - chip->gpiodev->base, chip->label);
+			gdev = gpiod_to_gpio_device(gpio_to_desc(gpio_num));
+		if (gdev)
+			seq_printf(s, "%u:%s ",
+				   gpio_num - gpio_device_get_base(gdev),
+				   gpio_device_get_label(gdev));
 		else
 			seq_puts(s, "0:? ");
 #endif
-- 
2.40.1

[PATCH 3/3] pinctrl: don't include GPIOLIB private header

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio_to_desc() is declared in linux/gpio.h so there's no need to include
gpiolib.h directly.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
---
 drivers/pinctrl/core.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/pinctrl/core.c b/drivers/pinctrl/core.c
index 0b4d07aea387..e55959e1af5a 100644
--- a/drivers/pinctrl/core.c
+++ b/drivers/pinctrl/core.c
@@ -24,6 +24,7 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 
+#include <linux/gpio.h>
 #include <linux/gpio/driver.h>
 
 #include <linux/pinctrl/consumer.h>
@@ -31,10 +32,6 @@
 #include <linux/pinctrl/machine.h>
 #include <linux/pinctrl/pinctrl.h>
 
-#ifdef CONFIG_GPIOLIB
-#include "../gpio/gpiolib.h"
-#endif
-
 #include "core.h"
 #include "devicetree.h"
 #include "pinconf.h"
-- 
2.40.1

Re: [PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Linus Walleij]

On Wed, Nov 15, 2023 at 5:50 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:

> From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
>
> Let's start working towards removing gpiod_to_chip() from the kernel.
> Start with pinctrl. The first patch should go through the GPIO tree
> and become available through an immutable tag so that we can get more
> conversions in this cycle elsewhere.

Fair enough, maybe an immutable branch with all three patches?
Probably best to merga all three into both subsystems I think.

Acked-by: Linus Walleij <linus.walleij@linaro.org>

Yours,
Linus Walleij

Re: [PATCH 3/3] pinctrl: don't include GPIOLIB private header

[Andy Shevchenko]

On Wed, Nov 15, 2023 at 05:50:01PM +0100, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> 
> gpio_to_desc() is declared in linux/gpio.h so there's no need to include
> gpiolib.h directly.

Oh, this is unfortunate we need it...

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Bartosz Golaszewski]

On Thu, Nov 16, 2023 at 2:43 PM Linus Walleij <linus.walleij@linaro.org> wrote:
>
> On Wed, Nov 15, 2023 at 5:50 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
>
> > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> >
> > Let's start working towards removing gpiod_to_chip() from the kernel.
> > Start with pinctrl. The first patch should go through the GPIO tree
> > and become available through an immutable tag so that we can get more
> > conversions in this cycle elsewhere.
>
> Fair enough, maybe an immutable branch with all three patches?
> Probably best to merga all three into both subsystems I think.
>

Nah, I don't think I need to pollute the GPIO tree with every pinctrl
patch. It's enough to just merge the first one into GPIO and you do
the same in pinctrl.

Bart

> Acked-by: Linus Walleij <linus.walleij@linaro.org>
>
> Yours,
> Linus Walleij

Re: [PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Linus Walleij]

On Mon, Nov 20, 2023 at 4:06 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> On Thu, Nov 16, 2023 at 2:43 PM Linus Walleij <linus.walleij@linaro.org> wrote:
> > On Wed, Nov 15, 2023 at 5:50 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> >
> > > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> > >
> > > Let's start working towards removing gpiod_to_chip() from the kernel.
> > > Start with pinctrl. The first patch should go through the GPIO tree
> > > and become available through an immutable tag so that we can get more
> > > conversions in this cycle elsewhere.
> >
> > Fair enough, maybe an immutable branch with all three patches?
> > Probably best to merga all three into both subsystems I think.
> >
>
> Nah, I don't think I need to pollute the GPIO tree with every pinctrl
> patch. It's enough to just merge the first one into GPIO and you do
> the same in pinctrl.

Fair enough, but I need that immutable branch for patch 1 so I
can merge the rest on top.

Yours,
Linus Walleij

Re: [PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Bartosz Golaszewski]

On Fri, Nov 24, 2023 at 11:01 AM Linus Walleij <linus.walleij@linaro.org> wrote:
>
> On Mon, Nov 20, 2023 at 4:06 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> > On Thu, Nov 16, 2023 at 2:43 PM Linus Walleij <linus.walleij@linaro.org> wrote:
> > > On Wed, Nov 15, 2023 at 5:50 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> > >
> > > > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> > > >
> > > > Let's start working towards removing gpiod_to_chip() from the kernel.
> > > > Start with pinctrl. The first patch should go through the GPIO tree
> > > > and become available through an immutable tag so that we can get more
> > > > conversions in this cycle elsewhere.
> > >
> > > Fair enough, maybe an immutable branch with all three patches?
> > > Probably best to merga all three into both subsystems I think.
> > >
> >
> > Nah, I don't think I need to pollute the GPIO tree with every pinctrl
> > patch. It's enough to just merge the first one into GPIO and you do
> > the same in pinctrl.
>
> Fair enough, but I need that immutable branch for patch 1 so I
> can merge the rest on top.
>
> Yours,
> Linus Walleij

I applied the first patch (after fixing the typo in the commit
message) and sent you the immutable branch to pull. Please apply the
remaining patches to the pinctrl tree directly.

Thanks!
Bart

Re: [PATCH 0/3] pinctrl: don't use gpiod_to_chip()

[Linus Walleij]

On Fri, Nov 24, 2023 at 8:40 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> On Fri, Nov 24, 2023 at 11:01 AM Linus Walleij <linus.walleij@linaro.org> wrote:
> >
> > On Mon, Nov 20, 2023 at 4:06 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> > > On Thu, Nov 16, 2023 at 2:43 PM Linus Walleij <linus.walleij@linaro.org> wrote:
> > > > On Wed, Nov 15, 2023 at 5:50 PM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
> > > >
> > > > > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> > > > >
> > > > > Let's start working towards removing gpiod_to_chip() from the kernel.
> > > > > Start with pinctrl. The first patch should go through the GPIO tree
> > > > > and become available through an immutable tag so that we can get more
> > > > > conversions in this cycle elsewhere.
> > > >
> > > > Fair enough, maybe an immutable branch with all three patches?
> > > > Probably best to merga all three into both subsystems I think.
> > > >
> > >
> > > Nah, I don't think I need to pollute the GPIO tree with every pinctrl
> > > patch. It's enough to just merge the first one into GPIO and you do
> > > the same in pinctrl.
> >
> > Fair enough, but I need that immutable branch for patch 1 so I
> > can merge the rest on top.
> >
> > Yours,
> > Linus Walleij
>
> I applied the first patch (after fixing the typo in the commit
> message) and sent you the immutable branch to pull. Please apply the
> remaining patches to the pinctrl tree directly.

Pulled it and applied the other two patches on top, thanks!

Yours,
Linus Walleij