I have been looking the ax25 source code, and didn't find nothing wrong.
So I tried to atenuate the problem of packets acumation disabling the feature
of "always poll the last frame of the tx windows", and that works. I still see
some repeating, but not the positive feedback wich caused that our gateway was
transmiting _very_much_ time the same frames without releasing the ptt (so
now there is surely a good reasing for such repeatings, but no more the "snow
ball" efect :-)
He I attach the patch, it's really easy:
cd [directorio_fuentes_linux]/net/ax25
cp ax25_out.c ax25_out.c.original
patch < [the_patch_here_attached]
And compile (if modules, only recompile modules)
In fact, ax25 v2.2 sdl diagrams say that the last frame must not be a polling
frame, and ax25 2.0 does not say about using p bit that way. As the ax25 states
are based in this sdl (or similar) diagrams, that can be the cause of
"something strange happens".
I think that this was done at the beginning, because timers were not
implemented, but now there is no reason for doing poll in the last frame of the
ax25 tx window; most of the implementation do not, and if that's causing
problems the best is remove it (imho).
--
Saludos de Juli�n
-.-
