On Sat, Jul 03, 2021 at 02:43:46AM +0800, kernel test robot wrote:
> 129cf89e585676 drivers/net/ethernet/intel/iavf/i40evf_ethtool.c Jesse Brandeburg 2018-09-14 1899 static int iavf_set_rxfh(struct net_device *netdev, const u32 *indir,
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1900 const u8 *key, const u8 hfunc)
> 4e9dc31f696ae8 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Mitch A Williams 2014-04-01 1901 {
> 129cf89e585676 drivers/net/ethernet/intel/iavf/i40evf_ethtool.c Jesse Brandeburg 2018-09-14 1902 struct iavf_adapter *adapter = netdev_priv(netdev);
> 2c86ac3c70794f drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Helin Zhang 2015-10-27 1903 u16 i;
> 4e9dc31f696ae8 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Mitch A Williams 2014-04-01 1904
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1905 /* We do not allow change in unsupported parameters */
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1906 if (key ||
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1907 (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP))
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1908 return -EOPNOTSUPP;
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1909 if (!indir)
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1910 return 0;
> 892311f66f2411 drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Eyal Perry 2014-12-02 1911
> 43a3d9ba34c9ca drivers/net/ethernet/intel/i40evf/i40evf_ethtool.c Mitch Williams 2016-04-12 @1912 memcpy(adapter->rss_key, key, adapter->rss_key_size);
Heh... There have been a bunch of patches modifying the behavior if
"key" is non-NULL and no one noticed that it's actually impossible.
:P
regards,
dan carpenter
