Re: [PATCH 1/1] dt-bindings: mfd: da9052: allow dlg,da9053-aa fallback to dlg,da9052

[Conor Dooley]

On Thu, Sep 03, 2026 at 12:04:42PM +0200, Krzysztof Kozlowski wrote:
> On Tue, Sep 01, 2026 at 06:00:31PM +0100, Conor Dooley wrote:
> > On Mon, Aug 31, 2026 at 04:15:21PM -0400, Frank.Li@xxxxxxxxxxx wrote:
> > > From: Frank Li <Frank.Li@xxxxxxx>
> > > 
> > > Allow dlg,da9053-aa fallback to dlg,da9052 to fix legacy platform (> 10
> > > years) check_dtbs warnings:
> > >   arch/arm/boot/dts/nxp/imx/imx53-voipac-bsb.dtb: dialog@48 (dlg,da9053-aa): compatible: ['dlg,da9053-aa', 'dlg,da9052'] is too long
> > >         from schema $id: http://devicetree.org/schemas/mfd/dlg,da9052.yaml
> > > 
> > > Signed-off-by: Frank Li <Frank.Li@xxxxxxx>
> > > ---
> > > Previous try drop dlg,da9052 at
> > > https://lore.kernel.org/imx/20260404-thoughtful-placid-boobook-da26da@quoll/
> > > 
> > > but Krzy have concern about compablity.
> > > 
> > > https://lore.kernel.org/imx/20251107-matrimony-showdown-ba6ce6996b2f@spud/
> > > 
> > > Conor prefer update old dts.
> > 
> > I don't think these devices are actually compatible, so this fallback
> > would be wrong.
> > The data comes from the match data, be that spi or i2c, and there's
> > explicitly different behaviour for the da9052 and da9053aa.
> 
> Maybe I looked at wrong driver, where is the difference? I made a very
> quick glance and I found the only difference in da9052_i2c_fix():
> 
>  59         switch (da9052->chip_id) {
>  60         case DA9052:
>  61         case DA9053_AA:
>  62         case DA9053_BA:
>  63         case DA9053_BB:
> 
> No differences in "da9052-core.c" and SPI, while looking for these
> device type enum values.

If you look further afield, since this is an mfd, in the regulator driver
there's a lot of conditions there that depend on the chip_id. E.g. current
limits and voltage steps differ. There's also one in the battery driver.

> > I doubt the legacy platform ever worked properly, and this is what should
> > have been explained in the PR that Krzysztof denied not just blindly
> > citing a dtbs_check problem.
> 
> Well, yeah, I wanted an accurate reason with actual analysis, not
> necessarily making devices compatible if they are not.
> 
> > 
> > Nacked-by: Conor Dooley <conor.dooley@xxxxxxxxxxxxx>
> 
> Best regards,
> Krzysztof
>

Attachment: signature.asc
Description: PGP signature