Change the behaviour of ip6_datagram_connect to consider the interface
set by the IPV6_UNICAST_IF socket option, similarly to udpv6_sendmsg.
This change is the IPv6 counterpart of the fix for IP_UNICAST_IF.
The tests introduced by that patch showed that the incorrect
behavior is present in IPv6 as well.
This patch fixes the broken test.
Reported-by: kernel test robot <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Fixes: 0e4d354762ce ("net-next: Fix IP_UNICAST_IF option behavior for connected sockets")
Signed-off-by: Richard Gobert <[email protected]>
---
net/ipv6/datagram.c | 15 ++++++++++-----
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index 7c7155b48f17..e624497fa992 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -42,24 +42,29 @@ static void ip6_datagram_flow_key_init(struct flowi6 *fl6, struct sock *sk)
{
struct inet_sock *inet = inet_sk(sk);
struct ipv6_pinfo *np = inet6_sk(sk);
+ int oif = sk->sk_bound_dev_if;
memset(fl6, 0, sizeof(*fl6));
fl6->flowi6_proto = sk->sk_protocol;
fl6->daddr = sk->sk_v6_daddr;
fl6->saddr = np->saddr;
- fl6->flowi6_oif = sk->sk_bound_dev_if;
fl6->flowi6_mark = sk->sk_mark;
fl6->fl6_dport = inet->inet_dport;
fl6->fl6_sport = inet->inet_sport;
fl6->flowlabel = np->flow_label;
fl6->flowi6_uid = sk->sk_uid;
- if (!fl6->flowi6_oif)
- fl6->flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
+ if (!oif)
+ oif = np->sticky_pktinfo.ipi6_ifindex;
- if (!fl6->flowi6_oif && ipv6_addr_is_multicast(&fl6->daddr))
- fl6->flowi6_oif = np->mcast_oif;
+ if (!oif) {
+ if (ipv6_addr_is_multicast(&fl6->daddr))
+ oif = np->mcast_oif;
+ else
+ oif = np->ucast_oif;
+ }
+ fl6->flowi6_oif = oif;
security_sk_classify_flow(sk, flowi6_to_flowi_common(fl6));
}
--
2.36.1
On 12/8/22 7:54 AM, Richard Gobert wrote:
> Change the behaviour of ip6_datagram_connect to consider the interface
> set by the IPV6_UNICAST_IF socket option, similarly to udpv6_sendmsg.
>
> This change is the IPv6 counterpart of the fix for IP_UNICAST_IF.
> The tests introduced by that patch showed that the incorrect
> behavior is present in IPv6 as well.
> This patch fixes the broken test.
>
> Reported-by: kernel test robot <[email protected]>
> Link: https://lore.kernel.org/r/[email protected]
> Fixes: 0e4d354762ce ("net-next: Fix IP_UNICAST_IF option behavior for connected sockets")
>
> Signed-off-by: Richard Gobert <[email protected]>
> ---
> net/ipv6/datagram.c | 15 ++++++++++-----
> 1 file changed, 10 insertions(+), 5 deletions(-)
>
Reviewed-by: David Ahern <[email protected]>
Hello:
This patch was applied to netdev/net.git (master)
by David S. Miller <[email protected]>:
On Thu, 8 Dec 2022 15:54:46 +0100 you wrote:
> Change the behaviour of ip6_datagram_connect to consider the interface
> set by the IPV6_UNICAST_IF socket option, similarly to udpv6_sendmsg.
>
> This change is the IPv6 counterpart of the fix for IP_UNICAST_IF.
> The tests introduced by that patch showed that the incorrect
> behavior is present in IPv6 as well.
> This patch fixes the broken test.
>
> [...]
Here is the summary with links:
- [v2] net: setsockopt: fix IPV6_UNICAST_IF option for connected sockets
https://git.kernel.org/netdev/net/c/526682b458b1
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html