Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 26 Sep 2017 11:45:55 +0200
From:      Julian Kornberger <jk@kornberger.name>
To:        Sepherosa Ziehau <sepherosa@gmail.com>
Cc:        "freebsd-net@freebsd.org" <freebsd-net@freebsd.org>
Subject:   Re: igb(4) txcsum6 broken in few cases
Message-ID:  <1df7456b-06c6-935f-c399-4bbb9d33280d@kornberger.name>
In-Reply-To: <CAMOc5cxY7ykP1sr%2Bt4C-4PfphSYCNHr2iBciua6s1R0rANhhPA@mail.gmail.com>
References:  <4cdc0f02-0ec4-4f1a-f14a-d4eeef78cbe0@kornberger.name> <CAMOc5cxY7ykP1sr%2Bt4C-4PfphSYCNHr2iBciua6s1R0rANhhPA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi Sepherosa,

I attached a dump. I can provide bigger dumps, but not in public, 
because they might contain sensitive data.
The outgoing UDP packets are created by my fastd kernel module [1] that 
is a simple UDP tunneling protocol that encapsulates IP packets using 
M_PREPEND. Some code is similar to if_vxlan.c.

ICMPv4 echo requests inside of the tunnel lead to invalid checksums in > 
95% of the cases. But also other packets types are affected.

[1] https://github.com/digineo/fastd/blob/master/kmod/fastd.c

Regards,
Julian


On 26.09.2017 07:10, Sepherosa Ziehau wrote:
> It will be more than helpful, if you can provides tcpdumps.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1df7456b-06c6-935f-c399-4bbb9d33280d>