[PATCH v4 00/10] gpio: improve support for shared GPIOs

[Bartosz Golaszewski]

On Tue, Nov 18, 2025 at 12:16 PM Geert Uytterhoeven
<geert at linux-m68k.org> wrote:
>
> Hi Bartosz,
>
> On Wed, 12 Nov 2025 at 15:05, Bartosz Golaszewski <brgl at bgdev.pl> wrote:
> > Bjorn, Konrad: I should have Cc'ed you on v1 but I just went with what
> > came out of b4 --auto-to-cc. It only gave me arm-msm. :( Patch 7 from
> > this series however impacts Qualcomm platforms. It's a runtime dependency
> > of patches 8 and 9. Would you mind Acking it so that I can take it into
> > an immutable branch that I'll make available to Mark Brown for him to
> > take patches 8-10 through the ASoC and regulator trees for v6.19?
> >
> > Problem statement: GPIOs are implemented as a strictly exclusive
> > resource in the kernel but there are lots of platforms on which single
> > pin is shared by multiple devices which don't communicate so need some
> > way of properly sharing access to a GPIO. What we have now is the
> > GPIOD_FLAGS_BIT_NONEXCLUSIVE flag which was introduced as a hack and
> > doesn't do any locking or arbitration of access - it literally just hand
> > the same GPIO descriptor to all interested users.
> >
> > The proposed solution is composed of three major parts: the high-level,
> > shared GPIO proxy driver that arbitrates access to the shared pin and
> > exposes a regular GPIO chip interface to consumers, a low-level shared
> > GPIOLIB module that scans firmware nodes and creates auxiliary devices
> > that attach to the proxy driver and finally a set of core GPIOLIB
> > changes that plug the former into the GPIO lookup path.
> >
> > The changes are implemented in a way that allows to seamlessly compile
> > out any code related to sharing GPIOs for systems that don't need it.
> >
> > The practical use-case for this are the powerdown GPIOs shared by
> > speakers on Qualcomm db845c platform, however I have also extensively
> > tested it using gpio-virtuser on arm64 qemu with various DT
> > configurations.
>
> Thanks for your series, part of which is now present linux-next.
> IIUIC, this requires the direction of the GPIO to be fixed?
>
> We have a long-standing use-case on various Renesas R-Car Gen3 boards
> (e.g. Salvator-X(S) and ULCB[1]), where GPIOs are shared by LEDs and
> key switches.  Basically, the GPIO is connected to:
>   1. A key switch connecting to GND when closed, with pull-up.
>   2. The gate of an N-channel MOSFET, turning on an LED when driven
>      high.
>
> Hence:
>   - In output mode, the LED can be controlled freely,
>   - In input mode, the LED is on, unless the key is pressed,
>   - Hence the switch state can only be read when the LED is turned on.
> If you have any idea how to handle this, feel free to reply ;-)
>
> Thanks!
>

How is this done currently? Even without this series and using the
GPIOD_FLAGS_BIT_NONEXCLUSIVE, the descriptor has a well-defined
direction so it's not possible for two drivers to request it as input
and output simultaneously. The second requester will override the
previous settings.

Bart