Re: [PATCH] dt-bindings: gpio: pca95xx: add onnn,pca9655 compatible

[Conor Dooley]

On Tue, Sep 08, 2026 at 12:37:21AM -0700, Bartosz Golaszewski wrote:
> On Mon, 7 Sep 2026 18:54:12 +0200, Conor Dooley <conor@xxxxxxxxxx> said:
> > On Fri, Sep 04, 2026 at 05:09:19PM -0700, Rosen Penev wrote:
> >> The pca9655 is the On Semi variant of the pca9555 16-bit I2C GPIO
> >> expander and is register-compatible with it. The addition of the
> >> compatible in the driver missed the documentation update.
> >
> > The commit message makes it sound like this device should use a
> > fallback. Why doesn't it? Either it should, or the commit message needs
> > an update.
> > pw-bot: changes-requested
> >
> 
> Actually commit 6d49b3a0f351 ("gpio: pca953x: Add support for the On Semi
> pca9655") added this compatible to the driver back in 2021 but it had remained
> undocumented until now. Unfortunately there's no denying I signed off on it. :(

Ye, but if patches like this appear for things that were undocumented
but ended up in drivers we still try to get the bindings correct when
they do show up. The commit message is basically telling me that this patch
is wrong and that a fallback to the pca9555 should be used. There's no
impact on the driver, so I don't see any harm in getting the binding
right.

> This commit should probably contain a Fixes: tag pointing at this commit.

Perhaps, but I'm not convinced that binding patches should ever have
Fixes tags citing drivers.

Attachment: signature.asc
Description: PGP signature