svn commit: r274709 - head/sys/netpfil/pf

[Ermal Luçi]

On Tue, Nov 25, 2014 at 3:14 PM, Dag-Erling Smørgrav <des at des.no> wrote:

> Ermal Luçi <eri at freebsd.org> writes:
> > Also this only affects the traffic sourced by the host itself and not
> > forwarded traffic and I think this patch will provide a regression for
> > the issues that the committed patch does.
>
> How?
>
The code as it stands (after your commit) is incorrect and will trigger
> an assertion in vtnet(4).  You could argue that it is less incorrect
> than the original, but the cure is worse than the disease.
>
>
Let me come back to you in the following days with a better analysis.
The existing sate of the world i think there will be checksum issues again
especially in policy routing situations.
I will come back with another patch to solve this, but am busy with some
${WORK} tasks.


> My patch fixes the panic as well as two preexisting bugs (not taking the
> IP checksum into account in the IPv4 path, and ignoring hardware
> offloading).  See https://bugs.freebsd.org/192013#c10 for an explanation
> of what it does and why.
>
> DES
> --
> Dag-Erling Smørgrav - des at des.no
>



-- 
Ermal