On Sat, Apr 05, 2014 at 10:48:36AM -0700, Thomas Pedersen wrote: > >> Does that mean we can get rid of net_traversal_jiffies entirely then? > >> PREQs are already gated by dot11MeshHWMPpreqMinInterval, so it seems > >> kind of redundant now. > > > > No, they are used correctly in limiting the increment rate of a STA's > > origin SN. I think we still need them there. > > Hmm. Ok then I'm missing something. Why is origin SN even incremented > when receiving a PREQ for which ifmsh is the target?
Now I'm confused: where do we do that? As I read it, Colleen's patch just aims to remove the arbitrary net_traversal_jiffies() on the PREQ case, and instead keep the same target sn until orig sn changes (i.e. orig sn is used to determine "is this a new path request?"). But orig sn should only change by the originator when another mesh_path_start_discovery() happens, right? -- Bob Copeland %% www.bobcopeland.com _______________________________________________ Devel mailing list [email protected] http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel
