On 9/13/24 06:28, Andrew Lunn wrote:
+static int fbnic_hwmon_sensor_id(enum hwmon_sensor_types type) +{ + if (type == hwmon_temp) + return FBNIC_SENSOR_TEMP; + if (type == hwmon_in) + return FBNIC_SENSOR_VOLTAGE; + + return -EOPNOTSUPP; +}+static int fbnic_hwmon_read(struct device *dev, enum hwmon_sensor_types type, + u32 attr, int channel, long *val) +{ + struct fbnic_dev *fbd = dev_get_drvdata(dev); + const struct fbnic_mac *mac = fbd->mac; + int id; + + id = fbnic_hwmon_sensor_id(type); + if (id < 0) + return -EOPNOTSUPP;fbnic_hwmon_sensor_id() itself returns EOPNOTSUPP, so just use it.+void fbnic_hwmon_register(struct fbnic_dev *fbd) +{ + if (!IS_REACHABLE(CONFIG_HWMON)) + return; + + fbd->hwmon = hwmon_device_register_with_info(fbd->dev, "fbnic", + fbd, &fbnic_chip_info, + NULL); + if (IS_ERR(fbd->hwmon)) { + dev_err(fbd->dev, + "Cannot register hwmon device %pe, aborting\n", + fbd->hwmon);aborting is probably the wrong word, because you keep going independent of it working or not.
I have not seen the original patch, and it doesn't seem to be available on the networking mailing list, so I can not really comment on the patch as a whole. For hwmon drivers in drivers/hwmon, I don't accept patches with dev_err() which is then ignored ... because ignoring the error means that it is not really an error. This should be dev_warn() or better dev_notice(). Guenter
