Re: [PATCH] ipvlan: fix ipv6 outbound device

That's a good question.
Recently, I am researching how to use ipvlan, I found when using L3 mode, the packet from
the ipvlan slave interface can only get out from the phy device which the slave interface attached,
and I found this is assigning the flowi4's output device result, so I tried to don't assign the
output device, I found the packet from the ipvlan slave interface can be forwarded to another
phy device, but it has bringing some other problem.
At the same time, when I read the code, I found the process of ipv6 packet is different, so I think
this is a bug.
I am look forward to listening to your advice!

---------------------------------------------------------------------------------------------------------------------------
On Thu, Nov 9, 2017 at 9:09 PM,  [off-list ref] wrote:

From: Keefe Liu <redacted>

When process the outbound packet of ipv6, we should assign the master 
device to output device other than input device.

curious to know, how you discovered this?

Signed-off-by: Keefe Liu <redacted>

Acked-by: Mahesh Bandewar <redacted>

quoted hunk

---
 drivers/net/ipvlan/ipvlan_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ipvlan/ipvlan_core.c 

b/drivers/net/ipvlan/ipvlan_core.c
index 034ae4c..f2a7e92 100644

--- a/drivers/net/ipvlan/ipvlan_core.c
+++ b/drivers/net/ipvlan/ipvlan_core.c

@@ -409,7 +409,7 @@ static int ipvlan_process_v6_outbound(struct sk_buff *skb)
        struct dst_entry *dst;
        int err, ret = NET_XMIT_DROP;
        struct flowi6 fl6 = {
-               .flowi6_iif = dev->ifindex,
+               .flowi6_oif = dev->ifindex,
                .daddr = ip6h->daddr,
                .saddr = ip6h->saddr,
                .flowi6_flags = FLOWI_FLAG_ANYSRC,
--

1.8.3.1