Jesse, don't we have the same bug in the datapath? I don't see any
special case for a zero UDP checksum there either.
Ethan, your fix looks fine, thanks.
On Wed, Dec 21, 2011 at 11:27:59AM -0800, Ethan Jackson wrote:
> dpif-netdev incorrectly calculated the UDP checksum when modifying
> source and destination ports.
>
> Signed-off-by: Ethan Jackson <et...@nicira.com>
> ---
> lib/dpif-netdev.c | 20 +++++++++++++++-----
> 1 files changed, 15 insertions(+), 5 deletions(-)
>
> diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
> index 2c416e3..742d00c 100644
> --- a/lib/dpif-netdev.c
> +++ b/lib/dpif-netdev.c
> @@ -1137,11 +1137,21 @@ dp_netdev_set_udp_port(struct ofpbuf *packet, const
> struct ovs_key_udp *udp_key)
> {
> struct udp_header *uh = packet->l4;
>
> - if (uh->udp_src != udp_key->udp_src) {
> - dp_netdev_set_port(&uh->udp_src, udp_key->udp_src, &uh->udp_csum);
> - }
> - if (uh->udp_dst != udp_key->udp_dst) {
> - dp_netdev_set_port(&uh->udp_dst, udp_key->udp_dst, &uh->udp_csum);
> + if (uh->udp_csum) {
> + if (uh->udp_src != udp_key->udp_src) {
> + dp_netdev_set_port(&uh->udp_src, udp_key->udp_src,
> &uh->udp_csum);
> + }
> +
> + if (uh->udp_dst != udp_key->udp_dst) {
> + dp_netdev_set_port(&uh->udp_dst, udp_key->udp_dst,
> &uh->udp_csum);
> + }
> +
> + if (!uh->udp_csum) {
> + uh->udp_csum = htons(0xffff);
> + }
> + } else {
> + uh->udp_src = udp_key->udp_src;
> + uh->udp_dst = udp_key->udp_dst;
> }
> }
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
