> +static int pse_set_pse_config(struct net_device *dev, > + struct netlink_ext_ack *extack, > + struct nlattr **tb) > +{ > + struct phy_device *phydev = dev->phydev; > + struct pse_control_config config = {}; > + int ret; > + > + if (!tb[ETHTOOL_A_PODL_PSE_ADMIN_CONTROL]) > + return -EINVAL; I would make use of extack here, and report what is missing. > + > + config.admin_cotrol = nla_get_u8(tb[ETHTOOL_A_PODL_PSE_ADMIN_CONTROL]); It would be good to have some basic validation here, make sure user space has passed a reasonable value. You should also define what 0 and ETHTOOL_A_PODL_PSE_ADMIN_CONTROL_UNKNOWN means here when passed in. In future, there could be additional things which could be configured, so struct pse_control_config gets additional members. ETHTOOL_A_PODL_PSE_ADMIN_CONTROL appears to be mandatory, you return -EVINAL if missing, so if you don't want to change it, but change some other new thing, maybe 0 should be passed here? And the driver should not consider it an error? ETHTOOL_A_PODL_PSE_ADMIN_CONTROL_UNKNOWN however seems invalid and so should be rejected here? Andrew