Hi Felix
We're recently come up against a problem whereby when you switch protocol to 3G/PPPoE, and then switch back to DHCP on the WAN interface, netifd is failing to bring up the WAN interface. I've tracked it down to a line inside /lib/netifd/dhcp.script: proto_init_update "*" 1 The problem here is that the * means that the interface name is not properly set, so for some reason it does not add the ip address, netmask, gateway or route to the correct WAN interface. If I manually set the interface name in that line, e.g.: proto_init_update "eth1" 1 All works fine. Can you advise what the * is used for, and why it's not the real interface name? Also, this problem only occurs when going back to DHCP from a 3G/PPPoE connection (that successfully dialled up and connected), so maybe something that happens during the PPP connection is stopping the * from being usable when you switch back to DHCP? For the moment I am using this to get around the problem: local realif=$(uci get network.wan.ifname) proto_init_update "$realif" 1 But I don't like hardcoding a UCI call in there when it should work by default. Any idea why this is happening? Thank you very much James Wood
_______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
