On Wed, 19 Mar 2025, Colin Foster wrote:
> On Wed, Mar 19, 2025 at 01:30:51PM +0100, Rasmus Villemoes wrote:
> > diff --git a/drivers/mfd/ocelot-core.c b/drivers/mfd/ocelot-core.c
> > index 41aff27088548..78b5fe15efdd2 100644
> > --- a/drivers/mfd/ocelot-core.c
> > +++ b/drivers/mfd/ocelot-core.c
> > @@ -200,10 +200,12 @@ static const struct mfd_cell vsc7512_devs[] = {
> > static void ocelot_core_try_add_regmap(struct device *dev,
> > const struct resource *res)
> > {
> > + struct ocelot_ddata *ddata = dev_get_drvdata(dev);
> > +
> > if (dev_get_regmap(dev, res->name))
> > return;
> >
> > - ocelot_spi_init_regmap(dev, res);
> > + ddata->init_regmap(dev, res);
>
> I remember changing this from function pointers to the direct function
> call during initial development, per Lee's suggestion. I like it though,
> and I'm glad to see multiple users now.
Yeah, we're still not going to be putting call-backs into device data.
Either pass the differentiating config through to the core driver or
handle the differentiation inside the *-i2c.c / *-spi.c files.
--
Lee Jones [李琼斯]
