[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MirageOS-devel] mirage-tcpip Fin/Fin-Ack



On 15 Feb 2014, at 16:05, Hannes Mehnert <hannes@xxxxxxxxxxx> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA384
> 
> Hey,
> 
> first of all: thanks for your great work on mirage. It is an awesome
> project.
> 
> While writing some mirage code, I discovered that the TCP layer sends
> a duplicate ack after a fin (which is fin + ack):
> 
> client --SYN-(seq=0)-> mirage
> mirage --SYN-ACK-(seq=0,ack=1)-> client
> client --ACK-(seq=1,ack=1)-> mirage
> client --data-(seq=1,len=7)-> mirage
> mirage --FIN-ACK-(seq=1,ack=8)-> client
> mirage --ACK-(seq=2,ack=8)-> client
> client --ACK-(seq=8,ack=2)-> mirage
> client --FIN-ACK-(seq=8,ack=2)-> mirage
> mirage --ACK-(seq=2,ack=9)-> client
> 
> 
> This happens already with the network-skeleton/network code, running
> on unix (FreeBSD) with a tap interface. I've uploaded a pcap file with
> the tcp connection to https://berlin.ccc.de/~hannes/mirage-network.pcap .
> 
> Unfortunately I wasn't able to spot the piece of code responsible for
> sending the duplicated ack.

I'll need to do a bit more digging here to pin this down, as the duplicate
ACK code has bounced between Balraj and me in recent months. 

-anil


_______________________________________________
MirageOS-devel mailing list
MirageOS-devel@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/mirageos-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.