On Mon, 5 Jul 2021 09:31:29 +0100
Jonathan Cameron [off-list ref] wrote:

On Mon, 5 Jul 2021 08:36:08 +0100
Lee Jones [off-list ref] wrote:

quoted

On Sat, 03 Jul 2021, Jonathan Cameron wrote:
  

quoted

On Sat,  3 Jul 2021 10:42:22 +0200
Andreas Kemnade [off-list ref] wrote:
    

quoted

This allows having devicetree nodes for the subdevices.

Signed-off-by: Andreas Kemnade <andreas@kemnade.info>
---
 drivers/mfd/rn5t618.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/mfd/rn5t618.c b/drivers/mfd/rn5t618.c
index 384acb459427..b916c7471ca3 100644
--- a/drivers/mfd/rn5t618.c
+++ b/drivers/mfd/rn5t618.c

@@ -24,8 +24,10 @@ static const struct mfd_cell rn5t618_cells[] = {
 };
 
 static const struct mfd_cell rc5t619_cells[] = {
-	{ .name = "rn5t618-adc" },
-	{ .name = "rn5t618-power" },
+	{ .name = "rn5t618-adc",
+	  .of_compatible = "ricoh,rc5t619-adc" },    

Odd to have a name of 618 and a compatible of 619.  Why?
Definitely deserves a comment if this is necessary for some reason!    

Actually this is the norm.  We have lots of drivers named after the
*first* device they supported before expansion.  

Ah. I'd missed that this cells array is specific to the 5t619, though if
the driver is the same I'd also expect it to be needed for the 5t618 entry.

Well, yes, it is needed for the 5t618 also. But if I would add it, it
would be untested. And that second shorter array is also used for the
rn5t567 which does not have an ADC, So I we need three arrays there.

Regards,
Andreas