Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-19 Thread Simon Wunderlich
Whoops, this is definitly a bug, thanks for pointing that out. Committed in r1552. On Mon, Jan 18, 2010 at 07:31:22PM +0800, Marek Lindner wrote: Hi, i've just committed this patch in revision 1551, as no further comments and critiques came in. I would consider this patch rather

Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-19 Thread Marek Lindner
On Monday 18 January 2010 23:24:26 Andrew Lunn wrote: gw_node_update() is missing a rcu_read_unlock(); before the return inside the list_for_each_entry_rcu() loop. The others are less serious, but should be corrected as well. Thanks - I just fixed it. What does this error message want to tell

Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-19 Thread Andrew Lunn
On Tue, Jan 19, 2010 at 08:12:49AM +0800, Marek Lindner wrote: On Monday 18 January 2010 23:24:26 Andrew Lunn wrote: gw_node_update() is missing a rcu_read_unlock(); before the return inside the list_for_each_entry_rcu() loop. The others are less serious, but should be corrected as well.

Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-18 Thread Andrew Lunn
On Mon, Jan 18, 2010 at 07:31:22PM +0800, Marek Lindner wrote: Hi, i've just committed this patch in revision 1551, as no further comments and critiques came in. I would consider this patch rather simple, and as experimental feature it is switched off by default anyways ... :)

Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-18 Thread Andrew Lunn
CHECK /home/lunn/batman/trunk/batman-adv-kernelland/gateway_common.c /home/lunn/batman/trunk/batman-adv-kernelland/gateway_common.c:28:6: warning: symbol 'kbit_to_gw_srv_class' was not declared. Should it be static? CHECK /home/lunn/batman/trunk/batman-adv-kernelland/gateway_client.c

Re: [B.A.T.M.A.N.] [PATCH] batman-adv: Add bonding functionality

2010-01-17 Thread Simon Wunderlich
Hey, i've just committed this patch in revision 1551, as no further comments and critiques came in. I would consider this patch rather simple, and as experimental feature it is switched off by default anyways ... :) I've changed only a small thing in the committed patch: When neighbors are