[lwip-users] [lwip] tcp on linux box sending reset

[vishwanath_kc]

Hello all,
Below is the tcpdump i took, TCP on server machine (linux RH 6.2) is
sending me reset
flag on, am unable to trace in either lwip or in linux source code.
Can anybody help me in fixing this up.

# tcpdump -i ppp0
17:22:49.452688 < 192.9.200.200.272 > 192.9.200.81.5376: S 6509:6509(0)
win 1024 <296>
17:22:49.452757 < 192.9.200.81.5376 > 192.9.200.200.272: R 0:0(0) ack
6510 win 0





> This is a multi-part message in MIME format.
> 
> 
> I had sent a mail yesterday describing the problems I was having when
running a TCP client on the LwIP side and a TCP server side on Linux
stack talking over a TAP/TUN interface.
> 
> I studied that problem some more. I think the latter problem (TCP
enqueue failure) was because of the previous problem of ARP failure.
When I ran tcpdump I noticed that the ARP packets sent by the LwIP side
had bad addresses in it. When I traced the functions I noticed the
comment in arp_query() function in arp.c saying "This is not
implemented".
> 
> So is ARP request from the LwIP side not implemented? Is that the
reason why I see bad addresses in the ARP request when I run tcpdump? Is
there some way to make this work?
> 
> Thanks,
> 
> Rajaram
> 
> 
> 
Are you iNabled
http://www.inablers.com
[This message was sent through the lwip discussion list.]