On Wed, Jul 11, 2018 at 05:19:15PM +0300, Dan Carpenter wrote:
> Hello Sergio Paracuellos,
>
> The patch 8ce76bff0e6a: "staging: ks7010: add new helpers to achieve
> mib set request and simplify code" from May 4, 2018, leads to the
> following static checker warning:
>
> drivers/staging/ks7010/ks_hostif.c:1845 hostif_sme_multicast_set()
> warn: '2' cannot fit into 'bool'
>
> drivers/staging/ks7010/ks_hostif.c
> 1831 static
> 1832 void hostif_sme_multicast_set(struct ks_wlan_private *priv)
> 1833 {
> 1834 struct net_device *dev = priv->net_dev;
> 1835 int mc_count;
> 1836 struct netdev_hw_addr *ha;
> 1837 char set_address[NIC_MAX_MCAST_LIST * ETH_ALEN];
> 1838 int i = 0;
> 1839
> 1840 spin_lock(&priv->multicast_spin);
> 1841
> 1842 memset(set_address, 0, NIC_MAX_MCAST_LIST * ETH_ALEN);
> 1843
> 1844 if (dev->flags & IFF_PROMISC) {
> 1845 hostif_mib_set_request_bool(priv, LOCAL_MULTICAST_FILTER,
> 1846 MCAST_FILTER_PROMISC);
> ^^^^^^^^^^^^^^^^^^^^^
> This is 2 but casted to MCAST_FILTER_MCASTALL because hostif_mib_set_request_bool() takes a
> bool.
Yes, you are totally right. This seems to be wrong. We should pass an int instead
of a bool to avoid this problem.
>
> In the original code, we used to pass 2 but as MIB_VALUE_TYPE_BOOL type.
> That's almost certainly buggy, but I don't know if it used to work on
> not.
Me neither but it is clear this is not what th original code was doing.
>
> 1847 goto spin_unlock;
> 1848 }
> 1849
> 1850 if ((netdev_mc_count(dev) > NIC_MAX_MCAST_LIST) ||
> 1851 (dev->flags & IFF_ALLMULTI)) {
> 1852 hostif_mib_set_request_bool(priv, LOCAL_MULTICAST_FILTER,
> 1853 MCAST_FILTER_MCASTALL);
> 1854 goto spin_unlock;
> 1855 }
>
> regards,
> dan carpenter
Best regards,
Sergio Paracuellos
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
