diff mbox series

pinctrl: bcm2835: Make pin freeing behavior configurable

Message ID 20240419204057.86078-1-wahrenst@gmx.net (mailing list archive)
State New, archived
Headers show
Series pinctrl: bcm2835: Make pin freeing behavior configurable | expand

Commit Message

Stefan Wahren April 19, 2024, 8:40 p.m. UTC 
Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
So in case it was configured as GPIO_OUT before the configured output
level also get lost. As long as GPIO sysfs was used this wasn't
actually a problem because the pins and their possible output level
were kept by sysfs.

Since more and more Raspberry Pi users start using libgpiod they are
confused about this behavior. So make the pin freeing behavior of
GPIO_OUT configurable via module parameter. In case
pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.

This patch based on the downstream work of Phil Elwell.

Link: https://github.com/raspberrypi/linux/pull/6117
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
---
 drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)

--
2.34.1

Comments

Stefan Wahren May 2, 2024, 10:22 a.m. UTC | #1 
Am 19.04.24 um 22:40 schrieb Stefan Wahren:
> Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
> So in case it was configured as GPIO_OUT before the configured output
> level also get lost. As long as GPIO sysfs was used this wasn't
> actually a problem because the pins and their possible output level
> were kept by sysfs.
>
> Since more and more Raspberry Pi users start using libgpiod they are
> confused about this behavior. So make the pin freeing behavior of
> GPIO_OUT configurable via module parameter. In case
> pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.
>
> This patch based on the downstream work of Phil Elwell.
>
> Link: https://github.com/raspberrypi/linux/pull/6117
> Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
> ---
>   drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
>   1 file changed, 16 insertions(+), 4 deletions(-)
Gentle ping ...
Kent Gibson May 2, 2024, 11:04 a.m. UTC | #2 
On Thu, May 02, 2024 at 12:22:07PM +0200, Stefan Wahren wrote:
> Am 19.04.24 um 22:40 schrieb Stefan Wahren:
> > Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
> > So in case it was configured as GPIO_OUT before the configured output
> > level also get lost. As long as GPIO sysfs was used this wasn't
> > actually a problem because the pins and their possible output level
> > were kept by sysfs.
> >
> > Since more and more Raspberry Pi users start using libgpiod they are
> > confused about this behavior. So make the pin freeing behavior of
> > GPIO_OUT configurable via module parameter. In case
> > pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.
> >
> > This patch based on the downstream work of Phil Elwell.
> >
> > Link: https://github.com/raspberrypi/linux/pull/6117
> > Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
> > ---
> >   drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
> >   1 file changed, 16 insertions(+), 4 deletions(-)
> Gentle ping ...

I can't comment on the substance of the change as pinctrl is outside my
wheelhouse, but the "strict_gpiod" name could be better.
The point is to make GPIO outputs persist, right?
The name should better reflect that.

Cheers,
Kent.
Stefan Wahren May 2, 2024, 11:11 a.m. UTC | #3 
Hi Kent,

Am 02.05.24 um 13:04 schrieb Kent Gibson:
> On Thu, May 02, 2024 at 12:22:07PM +0200, Stefan Wahren wrote:
>> Am 19.04.24 um 22:40 schrieb Stefan Wahren:
>>> Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
>>> So in case it was configured as GPIO_OUT before the configured output
>>> level also get lost. As long as GPIO sysfs was used this wasn't
>>> actually a problem because the pins and their possible output level
>>> were kept by sysfs.
>>>
>>> Since more and more Raspberry Pi users start using libgpiod they are
>>> confused about this behavior. So make the pin freeing behavior of
>>> GPIO_OUT configurable via module parameter. In case
>>> pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.
>>>
>>> This patch based on the downstream work of Phil Elwell.
>>>
>>> Link: https://github.com/raspberrypi/linux/pull/6117
>>> Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
>>> ---
>>>    drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
>>>    1 file changed, 16 insertions(+), 4 deletions(-)
>> Gentle ping ...
> I can't comment on the substance of the change as pinctrl is outside my
> wheelhouse, but the "strict_gpiod" name could be better.
> The point is to make GPIO outputs persist, right?
Yes, correct.
> The name should better reflect that.
Finding good and short names is hard, do you have a suggestion?

Thanks
>
> Cheers,
> Kent.
>
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Kent Gibson May 2, 2024, 11:15 a.m. UTC | #4 
On Thu, May 02, 2024 at 01:11:06PM +0200, Stefan Wahren wrote:
> Hi Kent,
>
> Am 02.05.24 um 13:04 schrieb Kent Gibson:
> > On Thu, May 02, 2024 at 12:22:07PM +0200, Stefan Wahren wrote:
> > > Am 19.04.24 um 22:40 schrieb Stefan Wahren:
> > > > Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
> > > > So in case it was configured as GPIO_OUT before the configured output
> > > > level also get lost. As long as GPIO sysfs was used this wasn't
> > > > actually a problem because the pins and their possible output level
> > > > were kept by sysfs.
> > > >
> > > > Since more and more Raspberry Pi users start using libgpiod they are
> > > > confused about this behavior. So make the pin freeing behavior of
> > > > GPIO_OUT configurable via module parameter. In case
> > > > pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.
> > > >
> > > > This patch based on the downstream work of Phil Elwell.
> > > >
> > > > Link: https://github.com/raspberrypi/linux/pull/6117
> > > > Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
> > > > ---
> > > >    drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
> > > >    1 file changed, 16 insertions(+), 4 deletions(-)
> > > Gentle ping ...
> > I can't comment on the substance of the change as pinctrl is outside my
> > wheelhouse, but the "strict_gpiod" name could be better.
> > The point is to make GPIO outputs persist, right?
> Yes, correct.
> > The name should better reflect that.
> Finding good and short names is hard, do you have a suggestion?
>

How about "persist_gpio_outputs"?

Cheers,
Kent.
Stefan Wahren May 2, 2024, 12:20 p.m. UTC | #5 
Am 02.05.24 um 13:15 schrieb Kent Gibson:
> On Thu, May 02, 2024 at 01:11:06PM +0200, Stefan Wahren wrote:
>> Hi Kent,
>>
>> Am 02.05.24 um 13:04 schrieb Kent Gibson:
>>> On Thu, May 02, 2024 at 12:22:07PM +0200, Stefan Wahren wrote:
>>>> Am 19.04.24 um 22:40 schrieb Stefan Wahren:
>>>>> Until now after a bcm2835 pin was freed its pinmux was set to GPIO_IN.
>>>>> So in case it was configured as GPIO_OUT before the configured output
>>>>> level also get lost. As long as GPIO sysfs was used this wasn't
>>>>> actually a problem because the pins and their possible output level
>>>>> were kept by sysfs.
>>>>>
>>>>> Since more and more Raspberry Pi users start using libgpiod they are
>>>>> confused about this behavior. So make the pin freeing behavior of
>>>>> GPIO_OUT configurable via module parameter. In case
>>>>> pinctrl-bcm2835.strict_gpiod is set to 0, the output level is kept.
>>>>>
>>>>> This patch based on the downstream work of Phil Elwell.
>>>>>
>>>>> Link: https://github.com/raspberrypi/linux/pull/6117
>>>>> Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
>>>>> ---
>>>>>     drivers/pinctrl/bcm/pinctrl-bcm2835.c | 20 ++++++++++++++++----
>>>>>     1 file changed, 16 insertions(+), 4 deletions(-)
>>>> Gentle ping ...
>>> I can't comment on the substance of the change as pinctrl is outside my
>>> wheelhouse, but the "strict_gpiod" name could be better.
>>> The point is to make GPIO outputs persist, right?
>> Yes, correct.
>>> The name should better reflect that.
>> Finding good and short names is hard, do you have a suggestion?
>>
> How about "persist_gpio_outputs"?
I'm fine with that. I'll wait until tomorrow before sending a new version.
>
> Cheers,
> Kent.
diff mbox series

Patch

diff --git a/drivers/pinctrl/bcm/pinctrl-bcm2835.c b/drivers/pinctrl/bcm/pinctrl-bcm2835.c
index f5a9372d43bd..a5b2096c97fc 100644
--- a/drivers/pinctrl/bcm/pinctrl-bcm2835.c
+++ b/drivers/pinctrl/bcm/pinctrl-bcm2835.c
@@ -244,6 +244,10 @@  static const char * const irq_type_names[] = {
 	[IRQ_TYPE_LEVEL_LOW] = "level-low",
 };

+static bool strict_gpiod = true;
+module_param(strict_gpiod, bool, 0644);
+MODULE_PARM_DESC(strict_gpiod, "unless true, GPIO_OUT remain when pin freed");
+
 static inline u32 bcm2835_gpio_rd(struct bcm2835_pinctrl *pc, unsigned reg)
 {
 	return readl(pc->base + reg);
@@ -926,6 +930,14 @@  static int bcm2835_pmx_free(struct pinctrl_dev *pctldev,
 		unsigned offset)
 {
 	struct bcm2835_pinctrl *pc = pinctrl_dev_get_drvdata(pctldev);
+	enum bcm2835_fsel fsel = bcm2835_pinctrl_fsel_get(pc, offset);
+
+	if (fsel == BCM2835_FSEL_GPIO_IN)
+		return 0;
+
+	/* preserve GPIO_OUT in non-strict mode */
+	if (!strict_gpiod && fsel == BCM2835_FSEL_GPIO_OUT)
+		return 0;

 	/* disable by setting to GPIO_IN */
 	bcm2835_pinctrl_fsel_set(pc, offset, BCM2835_FSEL_GPIO_IN);
@@ -970,10 +982,7 @@  static void bcm2835_pmx_gpio_disable_free(struct pinctrl_dev *pctldev,
 		struct pinctrl_gpio_range *range,
 		unsigned offset)
 {
-	struct bcm2835_pinctrl *pc = pinctrl_dev_get_drvdata(pctldev);
-
-	/* disable by setting to GPIO_IN */
-	bcm2835_pinctrl_fsel_set(pc, offset, BCM2835_FSEL_GPIO_IN);
+	bcm2835_pmx_free(pctldev, offset);
 }

 static int bcm2835_pmx_gpio_set_direction(struct pinctrl_dev *pctldev,
@@ -1419,6 +1428,9 @@  static int bcm2835_pinctrl_probe(struct platform_device *pdev)
 		goto out_remove;
 	}

+	dev_info(dev, "GPIO_OUT remain when pin freed: %s\n",
+		 strict_gpiod ? "no" : "yes");
+
 	return 0;

 out_remove: