[PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Arturas Moskvinas]

Chip outputs are enabled[1] before actual reset is performed[2] which might
cause pin output value to flip flop if previous pin value was set to 1 in chip.
Change fixes that behavior by making sure chip is fully reset before all outputs
are enabled.

Flip-flop can be noticed when module is removed and inserted again and one of
the pins was changed to 1 before removal. 100 microsecond flipping is
noticeable on oscilloscope (100khz SPI bus).

For a properly reset chip - output is enabled around 100 microseconds (on 100khz
SPI bus) later during probing process hence should be irrelevant behavioral
change.

[1] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L130
[2] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L150

Signed-off-by: Arturas Moskvinas <arturas.moskvinas@gmail.com>

---
v1 -> v2
* Updated commit message to contain more information why a change is made.
---
 drivers/gpio/gpio-74x164.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpio/gpio-74x164.c b/drivers/gpio/gpio-74x164.c
index e00c33310517..753e7be039e4 100644
--- a/drivers/gpio/gpio-74x164.c
+++ b/drivers/gpio/gpio-74x164.c
@@ -127,8 +127,6 @@ static int gen_74x164_probe(struct spi_device *spi)
 	if (IS_ERR(chip->gpiod_oe))
 		return PTR_ERR(chip->gpiod_oe);
 
-	gpiod_set_value_cansleep(chip->gpiod_oe, 1);
-
 	spi_set_drvdata(spi, chip);
 
 	chip->gpio_chip.label = spi->modalias;
@@ -153,6 +151,8 @@ static int gen_74x164_probe(struct spi_device *spi)
 		goto exit_destroy;
 	}
 
+	gpiod_set_value_cansleep(chip->gpiod_oe, 1);
+
 	ret = gpiochip_add_data(&chip->gpio_chip, chip);
 	if (!ret)
 		return 0;

base-commit: d206a76d7d2726f3b096037f2079ce0bd3ba329b
-- 
2.44.0

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Bartosz Golaszewski]

On Thu, Feb 29, 2024 at 9:47 AM Arturas Moskvinas
<arturas.moskvinas@gmail.com> wrote:
>
> Chip outputs are enabled[1] before actual reset is performed[2] which might
> cause pin output value to flip flop if previous pin value was set to 1 in chip.
> Change fixes that behavior by making sure chip is fully reset before all outputs
> are enabled.
>

Use imperative mood in commit messages - make it: "Fix that behavior
by making sure ...".

> Flip-flop can be noticed when module is removed and inserted again and one of
> the pins was changed to 1 before removal. 100 microsecond flipping is
> noticeable on oscilloscope (100khz SPI bus).
>
> For a properly reset chip - output is enabled around 100 microseconds (on 100khz
> SPI bus) later during probing process hence should be irrelevant behavioral
> change.
>
> [1] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L130
> [2] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L150
>
> Signed-off-by: Arturas Moskvinas <arturas.moskvinas@gmail.com>

This looks much better, can you add a Fixes: tag as well?

Bartosz

>
> ---
> v1 -> v2
> * Updated commit message to contain more information why a change is made.
> ---
>  drivers/gpio/gpio-74x164.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpio/gpio-74x164.c b/drivers/gpio/gpio-74x164.c
> index e00c33310517..753e7be039e4 100644
> --- a/drivers/gpio/gpio-74x164.c
> +++ b/drivers/gpio/gpio-74x164.c
> @@ -127,8 +127,6 @@ static int gen_74x164_probe(struct spi_device *spi)
>         if (IS_ERR(chip->gpiod_oe))
>                 return PTR_ERR(chip->gpiod_oe);
>
> -       gpiod_set_value_cansleep(chip->gpiod_oe, 1);
> -
>         spi_set_drvdata(spi, chip);
>
>         chip->gpio_chip.label = spi->modalias;
> @@ -153,6 +151,8 @@ static int gen_74x164_probe(struct spi_device *spi)
>                 goto exit_destroy;
>         }
>
> +       gpiod_set_value_cansleep(chip->gpiod_oe, 1);
> +
>         ret = gpiochip_add_data(&chip->gpio_chip, chip);
>         if (!ret)
>                 return 0;
>
> base-commit: d206a76d7d2726f3b096037f2079ce0bd3ba329b
> --
> 2.44.0
>

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Andy Shevchenko]

On Thu, Feb 29, 2024 at 10:45:56AM +0200, Arturas Moskvinas wrote:
> Chip outputs are enabled[1] before actual reset is performed[2] which might
> cause pin output value to flip flop if previous pin value was set to 1 in chip.
> Change fixes that behavior by making sure chip is fully reset before all outputs
> are enabled.
> 
> Flip-flop can be noticed when module is removed and inserted again and one of
> the pins was changed to 1 before removal. 100 microsecond flipping is
> noticeable on oscilloscope (100khz SPI bus).
> 
> For a properly reset chip - output is enabled around 100 microseconds (on 100khz
> SPI bus) later during probing process hence should be irrelevant behavioral
> change.

> [1] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L130
> [2] - https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L150

Please, convert these to be Link tags, so

at the end it will look like

Fixes:
Link: URL1 [1]
Link: URL2 [2]
Signed-off-by:

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Arturas Moskvinas]

On 2/29/24 10:54, Bartosz Golaszewski wrote:
> On Thu, Feb 29, 2024 at 9:47 AM Arturas Moskvinas
> <arturas.moskvinas@gmail.com> wrote:
>> Chip outputs are enabled[1] before actual reset is performed[2] which might
>> cause pin output value to flip flop if previous pin value was set to 1 in chip.
>> Change fixes that behavior by making sure chip is fully reset before all outputs
>> are enabled.
>>
> Use imperative mood in commit messages - make it: "Fix that behavior
> by making sure ...".
Chip outputs are enabled[1] before actual reset is performed[2] which might
cause pin output value to flip flop if previous pin value was set to 1 
in chip.
Fix that behavior by making sure chip is fully reset before all outputs are
enabled.

Bartosz, does this look better?

Arturas Moskvinas

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Bartosz Golaszewski]

On Thu, Feb 29, 2024 at 1:23 PM Arturas Moskvinas
<arturas.moskvinas@gmail.com> wrote:
>
>
>
> On 2/29/24 10:54, Bartosz Golaszewski wrote:
> > On Thu, Feb 29, 2024 at 9:47 AM Arturas Moskvinas
> > <arturas.moskvinas@gmail.com> wrote:
> >> Chip outputs are enabled[1] before actual reset is performed[2] which might
> >> cause pin output value to flip flop if previous pin value was set to 1 in chip.
> >> Change fixes that behavior by making sure chip is fully reset before all outputs
> >> are enabled.
> >>
> > Use imperative mood in commit messages - make it: "Fix that behavior
> > by making sure ...".
> Chip outputs are enabled[1] before actual reset is performed[2] which might
> cause pin output value to flip flop if previous pin value was set to 1
> in chip.
> Fix that behavior by making sure chip is fully reset before all outputs are
> enabled.
>
> Bartosz, does this look better?

... if previous pin value was set to 1 period

Bart

>
> Arturas Moskvinas

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Bartosz Golaszewski]

On Thu, Feb 29, 2024 at 6:24 PM Arturas Moskvinas
<arturas.moskvinas@gmail.com> wrote:
>
>
>
> On 2/29/24 15:25, Bartosz Golaszewski wrote:
>
> On Thu, Feb 29, 2024 at 1:23 PM Arturas Moskvinas
> <arturas.moskvinas@gmail.com> wrote:
>
> Chip outputs are enabled[1] before actual reset is performed[2] which might
> cause pin output value to flip flop if previous pin value was set to 1
> in chip.
> Fix that behavior by making sure chip is fully reset before all outputs are
> enabled.
>
> Bartosz, does this look better?
>
> ... if previous pin value was set to 1 period
>
> Andriy, Bartosz, is commit message ok?
> ===
> gpio: 74x164: Enable output pins after registers are reset
>
> Chip outputs are enabled[1] before actual reset is performed[2] which might
> cause pin output value to flip flop if previous pin value was set to 1.
> Fix that behavior by making sure chip is fully reset before all outputs are
> enabled.
>
> Flip-flop can be noticed when module is removed and inserted again and one of
> the pins was changed to 1 before removal. 100 microsecond flipping is
> noticeable on oscilloscope (100khz SPI bus).
>
> For a properly reset chip - output is enabled around 100 microseconds (on 100khz
> SPI bus) later during probing process hence should be irrelevant behavioral
> change.
>
> Fixes: 7ebc194d0fd4 (gpio: 74x164: Introduce 'enable-gpios' property)
> Link: https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L130 [1]
> Link: https://elixir.bootlin.com/linux/v6.7.4/source/drivers/gpio/gpio-74x164.c#L150 [2]
> Signed-off-by: Arturas Moskvinas <arturas.moskvinas@gmail.com>
> ===
>
> Arturas Moskvinas

Looks good to me.

Bart

Re: [PATCH v2] gpio: 74x164: Enable output pins after registers are reset

[Andy Shevchenko]

On Thu, Feb 29, 2024 at 07:24:09PM +0200, Arturas Moskvinas wrote:
> On 2/29/24 15:25, Bartosz Golaszewski wrote:
> > On Thu, Feb 29, 2024 at 1:23 PM Arturas Moskvinas
> > <arturas.moskvinas@gmail.com>  wrote:
> > > Chip outputs are enabled[1] before actual reset is performed[2] which might
> > > cause pin output value to flip flop if previous pin value was set to 1
> > > in chip.
> > > Fix that behavior by making sure chip is fully reset before all outputs are
> > > enabled.
> > > 
> > > Bartosz, does this look better?
> > ... if previous pin value was set to 1 period
> Andriy, Bartosz, is commit message ok?

LGTM! Please, send a v3 with it.

-- 
With Best Regards,
Andy Shevchenko