| Message ID | 20220713131421.1527179-4-nuno.sa@analog.com |
|---|---|
| State | New |
| Headers | show |
| Series | add support for bias pull-disable | expand |
On Wed, Jul 13, 2022 at 3:13 PM Nuno Sá <nuno.sa@analog.com> wrote: > On top of looking at PULL_UP and PULL_DOWN flags, also look at > PULL_DISABLE and set the appropriate GPIO flag. The GPIO core will then > pass down this to controllers that support it. > > Signed-off-by: Nuno Sá <nuno.sa@analog.com> Do we have a semantic check that PULLDOWN and PULLUP is not used in combination with NOPULL here? (We should also be checking that PULLDOWN and PULLUP are not used simultaneously but that is an unrelated thing.) Yours, Linus Walleij
On Mon, 2022-07-18 at 12:32 +0200, Linus Walleij wrote: > On Wed, Jul 13, 2022 at 3:13 PM Nuno Sá <nuno.sa@analog.com> wrote: > > > On top of looking at PULL_UP and PULL_DOWN flags, also look at > > PULL_DISABLE and set the appropriate GPIO flag. The GPIO core will > > then > > pass down this to controllers that support it. > > > > Signed-off-by: Nuno Sá <nuno.sa@analog.com> > > Do we have a semantic check that PULLDOWN and PULLUP > is not used in combination with NOPULL here? > > (We should also be checking that PULLDOWN and PULLUP > are not used simultaneously but that is an unrelated thing.) I did extended this check: https://elixir.bootlin.com/linux/v5.19-rc7/source/drivers/gpio/gpiolib.c#L3948 on patch 1 to make sure that PULLDOWN and PULLUP are not used with NOPULL. Is this what you have in mind or is it something else? - Nuno Sá
On Mon, Jul 18, 2022 at 12:48 PM Nuno Sá <noname.nuno@gmail.com> wrote: > On Mon, 2022-07-18 at 12:32 +0200, Linus Walleij wrote: > > On Wed, Jul 13, 2022 at 3:13 PM Nuno Sá <nuno.sa@analog.com> wrote: > > > > > On top of looking at PULL_UP and PULL_DOWN flags, also look at > > > PULL_DISABLE and set the appropriate GPIO flag. The GPIO core will > > > then > > > pass down this to controllers that support it. > > > > > > Signed-off-by: Nuno Sá <nuno.sa@analog.com> > > > > Do we have a semantic check that PULLDOWN and PULLUP > > is not used in combination with NOPULL here? > > > > (We should also be checking that PULLDOWN and PULLUP > > are not used simultaneously but that is an unrelated thing.) > > I did extended this check: > > https://elixir.bootlin.com/linux/v5.19-rc7/source/drivers/gpio/gpiolib.c#L3948 > > on patch 1 to make sure that PULLDOWN and PULLUP are not used with > NOPULL. Is this what you have in mind or is it something else? Excellent, thanks! That is exactly what we need, sorry for not being able to keep all floating patches in my head :D Yours, Linus Walleij
On Wed, Jul 13, 2022 at 03:14:20PM +0200, Nuno Sá wrote: > On top of looking at PULL_UP and PULL_DOWN flags, also look at > PULL_DISABLE and set the appropriate GPIO flag. The GPIO core will then > pass down this to controllers that support it. In case this patch will be in v2, Acked-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> P.S. I will be on vacation till mid-August.
diff --git a/drivers/gpio/gpiolib-acpi.c b/drivers/gpio/gpiolib-acpi.c index c2523ac26fac..9be1376f9a62 100644 --- a/drivers/gpio/gpiolib-acpi.c +++ b/drivers/gpio/gpiolib-acpi.c @@ -687,6 +687,9 @@ int acpi_gpio_update_gpiod_lookup_flags(unsigned long *lookupflags, case ACPI_PIN_CONFIG_PULLDOWN: *lookupflags |= GPIO_PULL_DOWN; break; + case ACPI_PIN_CONFIG_NOPULL: + *lookupflags |= GPIO_PULL_DISABLE; + break; default: break; }
On top of looking at PULL_UP and PULL_DOWN flags, also look at PULL_DISABLE and set the appropriate GPIO flag. The GPIO core will then pass down this to controllers that support it. Signed-off-by: Nuno Sá <nuno.sa@analog.com> --- drivers/gpio/gpiolib-acpi.c | 3 +++ 1 file changed, 3 insertions(+)