Hi Alexander, On Mon, Jul 25, 2022 at 08:38:57AM +0200, Alexander Stein wrote: > Hi Sakari, > > Am Freitag, 22. Juli 2022, 15:50:30 CEST schrieb Sakari Ailus: > > Hi Alexander, > > > > On Fri, Jul 22, 2022 at 03:19:46PM +0200, Alexander Stein wrote: > > > To distinguish ov9281 & ov9282 the name has to be explicitly set. > > > i2c_client already has the name parsed from the compatible. > > > > > > Signed-off-by: Alexander Stein <alexander.stein@xxxxxxxxxxxxxxx> > > > --- > > > > > > drivers/media/i2c/ov9282.c | 1 + > > > 1 file changed, 1 insertion(+) > > > > > > diff --git a/drivers/media/i2c/ov9282.c b/drivers/media/i2c/ov9282.c > > > index 352dbe21a902..dbc0a4cd060f 100644 > > > --- a/drivers/media/i2c/ov9282.c > > > +++ b/drivers/media/i2c/ov9282.c > > > @@ -1047,6 +1047,7 @@ static int ov9282_probe(struct i2c_client *client) > > > > > > /* Initialize subdev */ > > > v4l2_i2c_subdev_init(&ov9282->sd, client, &ov9282_subdev_ops); > > > > > > + v4l2_i2c_subdev_set_name(&ov9282->sd, client, client->name, NULL); > > > > Could you instead do this based on the compatible string in the driver, > > using device_get_match_data()? The approach works on non-OF systems, too. > > I actually don't like doing the same as of_modalias_node() is doing. > Until non-OF support is added (if ever), I don't see any benefit in doing so > right now. client->name will be wrong on un-OF; putting this string to device match data is a better option. It'll be about the same number of lines, too. -- Sakari Ailus
