Re: [PATCH 00/13] Input: adp5589: refactor and platform_data removal

["Nuno Sá" <noname.nuno@gmail.com>]

On Tue, 2024-10-01 at 15:41 +0200, Nuno Sa wrote:
> This series aims to remove platform data dependency from the adp5589
> driver (as no platform is really using it) and instead add support for
> FW properties. Note that rows and columns for the keypad are being given
> as masks and that was briefly discussed with Dmitry. For context
> on why this is being done as mask [1].
> 
> The first couple of patches are fixes that we may want to backport...
> 
> [1]:
> https://lore.kernel.org/linux-input/9db96c99c805e615ba40ca7fd3632174d1e8d11f.camel@gmail.com/
> 
> ---
> Nuno Sa (13):
>       Input: adp5589-keys: fix NULL pointer dereference
>       Input: adp5589-keys: fix adp5589_gpio_get_value()
>       Input: adp5589-keys: add chip_info structure
>       Input: adp5589-keys: support gpi key events as 'gpio keys'
>       dt-bindings: input: Document adp5589 and similar devices
>       Input: adp5589-keys: add support for fw properties
>       Input: adp5589-keys: add guard() notation
>       Input: adp5589-keys: bail out on returned error
>       Input: adp5589-keys: refactor adp5589_read()
>       Input: adp5589-keys: fix coding style
>       Input: adp5589-keys: unify adp_constants in info struct
>       Input: adp5589-keys: make use of dev_err_probe()
>       Input: adp5589-keys: add regulator support
> 
>  .../devicetree/bindings/input/adi,adp5589.yaml     |  310 +++++
>  .../devicetree/bindings/trivial-devices.yaml       |    6 -
>  MAINTAINERS                                        |    8 +
>  drivers/input/keyboard/Kconfig                     |    3 +
>  drivers/input/keyboard/adp5589-keys.c              | 1397 +++++++++++++-------
>  include/linux/input/adp5589.h                      |  180 ---
>  6 files changed, 1254 insertions(+), 650 deletions(-)
> ---
> base-commit: c7bf046925dc5885d9c4d8fbcbb7e4e73665bfcf
> change-id: 20240930-b4-dev-adp5589-fw-conversion-955b2f42da70
> --
> 
> Thanks!
> - Nuno Sá
> 

Hi Dmitry,

Something really caught my attention now while checking 6.12 merge window. It seems
we have a new MFD device for adp5585 [1] which adds duplicated functionality (that
was already present in adp5589-keys.c). So, having this as MFD might makes sense
(even though it makes it harder to validate the keys and to make use of gpio-keys)
but we are now duplicating GPIO support. Bottom line, not sure what we should do next
and should I proceed for v2?

Also ccing Lee and Bartosz...

[1]: https://lore.kernel.org/lkml/20240816152738.GB5853@google.com/

- Nuno Sá