Where are you capturing the network traffic? I've noticed some checksum weirdness in tcpdump with NICs that provide hardware checksum.
Are you seeing bad header checksums in netstat -s ? Skye Word on the street is that Dan Langille said: > I am forwarding this on behalf of Bruce (his DNS is borked and thus he > cannot send to the lists). Please don't CC me in the replies, only Bruce. > > Thanks. > > ------- Forwarded message follows ------- > Date sent: Fri, 25 Jan 2002 23:35:38 +0000 > From: Bruce M Simpson <[EMAIL PROTECTED]> > To: [EMAIL PROTECTED] > Subject: ip_output() does not checksum outer header > > Dan, > > if you could forward this on to freebsd-net I would be most grateful. > > I have completed most of the code to implement GRE as a regular > interface driver under 4.4-STABLE using the cloner interface. This > has all worked fine in local tests. Tonight, whilst testing with > a remote site, we noticed that the encapsulating IP datagram header > around the GRE header does not have any header checksum. > > I have checked, checked and re-checked the m->m_len, m->m_pkthdr.len, > m_pullup() results et al and can find nothing out of the ordinary. > > This is driving me mad right now, can anybody shed any light on the > problem? > > Regards, > Bruce. > > ------- End of forwarded message ------- > > -- > Dan Langille > The FreeBSD Diary - http://freebsddiary.org/ - practical examples > > > To Unsubscribe: send mail to [EMAIL PROTECTED] > with "unsubscribe freebsd-hackers" in the body of the message -- 1966 MGB - GHN3L Safety Fast! __,__\__ The MGB Experience www.mgbexperience.com (_o____o_) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
