Hi Kevin, Frankly speaking, resulting following conf should be only treated as misconfiguration. > interface=br0 > no-dhcp-interface=br0 > interface=br0
As for tomato, it's up to fw logic/general settings why dhcp is disabled via web ui, but tried ti be enabled by user custom settings. So, there's nothing wrong with dnsmasq to be fixed, instead just drop no-dhcp-interface line from being saved. If there's no dhcp-range, no dhcp will happen anyway. Best Regards, Vladislav Grishenko > -----Original Message----- > From: dnsmasq-discuss-boun...@lists.thekelleys.org.uk [mailto:dnsmasq- > discuss-boun...@lists.thekelleys.org.uk] On Behalf Of Kevin Darbyshire- > Bryant > Sent: Wednesday, March 27, 2013 11:41 PM > To: dnsmasq-discuss@lists.thekelleys.org.uk > Subject: [Dnsmasq-discuss] Config file behaviour - change/bug/feature - > comments > > Hi Simon, > > As previously mentioned I got 2.66test16 into a recent version of Tomato > router firmware which means it's now out in the wild and being used. An > obscure corner case change in behaviour has been observed, relating to > disabling and then re-enabling dhcp service on an interface, and specifically > which lines in a config file stick. Example config: > > interface=br0 > no-dhcp-interface=br0 > > then later in the file: > > interface=br0 > dhcp-range=br0,192.168.235.21,192.168.235.254,255.255.255.0,720m > dhcp-option=br0,3,192.168.235.1 > dhcp-option=br0,44,0.0.0.0 > > Now in v2.61 (!) dnsmasq would take the last interface line and say 'oh you > do want dns AND dhcp service' and everything would procede as expected. > In 2.66, once no-dhcp-interface is set it's set and dhcp service is disabled > forever. Now there's a very valid argument that the above example is a > stupid config file. Unfortunately the example config file is representative of > the conflicts between the automatically generated part of Tomato's dnsmasq > config, and the manual 'advanced config options' that get appended to the > file via the router's GUI. > > Now I personally think the router's gui should be fixed (though I've no idea > how at this point) but is there perhaps an argument that an 'interface=intfn' > command later in the file should also reset the 'no-dhcp-flag'? > > Kevin _______________________________________________ Dnsmasq-discuss mailing list Dnsmasq-discuss@lists.thekelleys.org.uk http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
