Skip to content

Commit

Permalink
ipv4: Don't use the cached pmtu informations for input routes
Browse files Browse the repository at this point in the history
The pmtu informations on the inetpeer are visible for output and
input routes. On packet forwarding, we might propagate a learned
pmtu to the sender. As we update the pmtu informations of the
inetpeer on demand, the original sender of the forwarded packets
might never notice when the pmtu to that inetpeer increases.
So use the mtu of the outgoing device on packet forwarding instead
of the pmtu to the final destination.

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
  • Loading branch information
Steffen Klassert authored and David S. Miller committed Nov 26, 2011
1 parent b8400f3 commit 261663b
Showing 1 changed file with 2 additions and 2 deletions.
4 changes: 2 additions & 2 deletions net/ipv4/route.c
Original file line number Diff line number Diff line change
Expand Up @@ -1816,15 +1816,15 @@ static unsigned int ipv4_default_advmss(const struct dst_entry *dst)

static unsigned int ipv4_mtu(const struct dst_entry *dst)
{
const struct rtable *rt = (const struct rtable *) dst;
unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);

if (mtu)
if (mtu && rt_is_output_route(rt))
return mtu;

mtu = dst->dev->mtu;

if (unlikely(dst_metric_locked(dst, RTAX_MTU))) {
const struct rtable *rt = (const struct rtable *) dst;

if (rt->rt_gateway != rt->rt_dst && mtu > 576)
mtu = 576;
Expand Down

0 comments on commit 261663b

Please sign in to comment.