ipv6: always prefer rt6i_gateway if present
authorJulian Anastasov <ja@ssi.bg>
Sun, 20 Oct 2013 12:43:03 +0000 (15:43 +0300)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 4 Nov 2013 12:31:05 +0000 (04:31 -0800)
commit208a6152b633a33722cc53bca47c46b79e88a2ad
tree533f347093b805cbd47d788eb2bd441fb1d412d1
parentb90cd7b9d0baab2e8176d9cca5f18a592ef16063
ipv6: always prefer rt6i_gateway if present

[ Upstream commit 96dc809514fb2328605198a0602b67554d8cce7b ]

In v3.9 6fd6ce2056de2709 ("ipv6: Do not depend on rt->n in
ip6_finish_output2()." changed the behaviour of ip6_finish_output2()
such that the recently introduced rt6_nexthop() is used
instead of an assigned neighbor.

As rt6_nexthop() prefers rt6i_gateway only for gatewayed
routes this causes a problem for users like IPVS, xt_TEE and
RAW(hdrincl) if they want to use different address for routing
compared to the destination address.

Another case is when redirect can create RTF_DYNAMIC
route without RTF_GATEWAY flag, we ignore the rt6i_gateway
in rt6_nexthop().

Fix the above problems by considering the rt6i_gateway if
present, so that traffic routed to address on local subnet is
not wrongly diverted to the destination address.

Thanks to Simon Horman and Phil Oester for spotting the
problematic commit.

Thanks to Hannes Frederic Sowa for his review and help in testing.

Reported-by: Phil Oester <kernel@linuxace.com>
Reported-by: Mark Brooks <mark@loadbalancer.org>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
include/net/ip6_route.h