#6211 BLOC Future : patch for support of per packet mesh ttl

[Zarro Boogs per Child]

#6211: patch for support of per packet mesh ttl
--------------------------+-------------------------------------------------
  Reporter:  mbletsas     |       Owner:  dwmw2                         
      Type:  enhancement  |      Status:  new                           
  Priority:  blocker      |   Milestone:  Future Release                
 Component:  wireless     |     Version:                                
Resolution:               |    Keywords:  libertas, ttl, mesh, Update.1?
  Verified:  0            |    Blocking:                                
 Blockedby:               |  
--------------------------+-------------------------------------------------

Comment(by jcardona):

 Replying to [comment:4 cscott]:
 > Has this patch ever been tested?

 The [https://cozybit1.dnsalias.org/~javier/patches/0001-Support-for-using-
 setsockpt-to-change-the-mesh-ttl.patch.old original patch] had been well
 tested.  I must admit that the updated patch was a bit half hearted ...

 > Some fundamental problems:
 > (...)

 I believe I've addressed all of them in this
 [https://cozybit1.dnsalias.org/~javier/patches/0001-libertas-Per-socket-
 mesh-ttl-via-setsockopt.patch revised patch], except...

 >  * If get_sock_ttl were ever to be called, it would perform a linear
 search through a list of socket options, which could be unpleasant.
 (thanks, wad)

 There is one entry on that list per each socket with non-default ttl.  If
 we expect the number of concurrent non-default sockets to be large, then
 yes, a faster lookup is required.

 > The basic idea is not completely unreasonable.

 That's the nicest thing someone has told me in ages :)

 > The trick here is that we want a link-layer ttl, not a IP ttl, since
 mesh hops aren't IP hops.

 I agree with that.  It's just that the hooks are there for IP and not for
 the link layer.

 > The suggestion at http://lists.openwall.net/netdev/2007/08/16/141 of
 using a generic SO_LL_TTL
 > socket option seems reasonable to me, since you can have link-layer hops
 even on a switched
 > ethernet network (say).

 Yes, we proposed that and if you follow that thread, you see that there
 was not much response.
 I'd like to have some assurance that this will be accepted upstream before
 going that route.
 You'll have to agree that the current implementation is minimally
 intrusive:  it only declares one new sock option, and yet it was rejected.
 Implementing a link layer option would require lot more changes to the
 core stack.

 > This patch, however, is certainly not ready to apply.

 Agreed.  But [https://cozybit1.dnsalias.org/~javier/patches/0001-libertas-
 Per-socket-mesh-ttl-via-setsockopt.patch this one] is.  As soon as we
 finish testing the new version we'll send that to libertas-dev.

-- 
Ticket URL: <http://dev.laptop.org/ticket/6211#comment:10>
One Laptop Per Child <http://dev.laptop.org>
OLPC bug tracking system