On Tue, 2012-09-18 at 13:45 +0200, Johannes Berg wrote: > On Tue, 2012-09-18 at 10:44 +0200, Johannes Berg wrote: > > > Thanks. I just got that far as well, but I hadn't set up a DHCP > > server ... I'll debug further, I just confirmed that association still > > works. > > Ok I can't reproduce the problem. You're going to have to be more > specific about what connman sets up, I think. > > Here's what I did: > > * brctl addbr br0 > * ip addr add ... br0 > * ip link set br0 up > * start dnsmasq on br0 > * start wpa_supplicant on wlan0, with -Dnl80211 -bbr0 -iwlan0 -c../cfg > * brctl addif br0 wlan0 > > Maybe this isn't exactly what connman does, but it works for me with > kernel 3.5.4, a 6230 Intel device and 1.0 wpa_supplicant. > > Can you capture exactly what connman does? I can with some amount of efforts - I'd need to read the code. If you would be kind / enthusiastic enough to try this with Tizen images, that would be a lot easier for me. But let me CC Samuel who did tethering support for connman. Samuel, if you wish to provide some input, here is the link to this discussion: http://thread.gmane.org/gmane.linux.kernel.wireless.general/97389 > Also, are you sure that reverting just the carrier state patch isn't > sufficient? It seems to me that the carrier state patch might confuse > the bridge code, but I really see no reason Eric's change should have > any impact at all. Yes, 100% sure. I spent a lot of time bisecting this, and found out that in my setup tethering starts working only if I revert both. -- Best Regards, Artem Bityutskiy
Attachment:
signature.asc
Description: This is a digitally signed message part