[Ns-developers] wifi AP behavior change

[Gustavo Carneiro]

2009/2/20 Gustavo Carneiro <gjcarneiro at gmail.com>

>
>
> 2009/2/19 Mathieu Lacage <mathieu.lacage at sophia.inria.fr>
>
> hi,
>>
>> While reading the nqap-wifi-mac.cc file, it occured to me that the
>> packet forwarding logic was wrong. The attached patch attempts to make
>> it slightly more correct and comments the code (I must be growing old).
>> This patch is going to change regression output, and, more generally,
>> changes the behavior of the code so, I would appreciate a review of this
>> patch and some testing. Gustavo, Kirillano, others ?
>
>
> Mathieu, I think the patch makes perfect sense.  The Learning Bridge
> attached to the wifi device, if any, does not need to be involved when a
> frame is being forwarded by the AP between to stations both associated to
> it.
>

On second thought, after applying the patch and testing wifi-wired-bridging,
seems do not seem to working as they should.  I see no unicast data flowing,
and I see ARP requests, but the ARP replies are not reaching the requesting
node.  Something is wrong.  Maybe not in the patch, but maybe the patch
uncovers some other bug.

In theory the patch looks nice, but in practice things may not be so
simple.  I would advise caution before comitting the patch.

-- 
Gustavo J. A. M. Carneiro
INESC Porto, Telecommunications and Multimedia Unit
"The universe is always one step beyond logic." -- Frank Herbert