On 20/04/23 04:06, Guenter Roeck wrote: > On Wed, Apr 19, 2023 at 11:36:56AM +1200, Chris Packham wrote: >> Instead of of_property_read_*() use the equivalent >> device_property_read_*() API. This will allow these properties to be >> used on DT unaware platforms. For DT aware platforms this will be a >> noop. >> >> Signed-off-by: Chris Packham <chris.packham@xxxxxxxxxxxxxxxxxxx> >> --- >> >> Notes: >> This is an additional update for master from the preceeding bugfix >> commit. I've not added a fixes tag for this one because I don't think >> there will be a behaviour change for existing usages. >> >> I know we have one upcoming DT unaware platform that we may want to use >> some of these properties via ACPI tables so I won't object if this ends >> up on the stable track but I don't think it meets the criteria for >> stable. >> >> drivers/hwmon/adt7475.c | 8 ++++---- >> 1 file changed, 4 insertions(+), 4 deletions(-) >> >> diff --git a/drivers/hwmon/adt7475.c b/drivers/hwmon/adt7475.c >> index 6a6ebcc896b1..3b9289bc5997 100644 >> --- a/drivers/hwmon/adt7475.c >> +++ b/drivers/hwmon/adt7475.c >> @@ -1468,7 +1468,7 @@ static int load_config3(const struct i2c_client *client, const char *propname) >> u8 config3; >> int ret; >> >> - ret = of_property_read_string(client->dev.of_node, propname, &function); >> + ret = device_property_read_string(&client->dev, propname, &function); > Unfortunately that is a problem because the parameter passed to > load_config3 is a pointer to a const data structure and > device_property_read_string doesn't like that (afaics for > no good reason). You'll also have to change the client parameter > to load_config() and friends to not be const. Not sure how I didn't notice that. Probably rushing to get the fix part out. I do have a v3 that drops the const where needed to silence the warnings but I'll sit on that for now in the hope that your patch gets accepted. > Guenter
