ip_tunnel: fix over-mtu packet send fail without TUNNEL_DONT_FRAGMENT…

… flags

[ Upstream commit 20149e9 ]

The tunnel device such as vxlan, bareudp and geneve in the lwt mode set
the outer df only based TUNNEL_DONT_FRAGMENT.
And this was also the behavior for gre device before switching to use
ip_md_tunnel_xmit in commit 962924f ("ip_gre: Refactor collect
metatdata mode tunnel xmit to ip_md_tunnel_xmit")

When the ip_gre in lwt mode xmit with ip_md_tunnel_xmi changed the rule and
make the discrepancy between handling of DF by different tunnels. So in the
ip_md_tunnel_xmit should follow the same rule like other tunnels.

Fixes: cfc7381 ("ip_tunnel: add collect_md mode to IPIP tunnel")
Signed-off-by: wenxu <wenxu@ucloud.cn>
Link: https://lore.kernel.org/r/1604028728-31100-1-git-send-email-wenxu@ucloud.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net/ipv4/ip_tunnel.c

@@ -614,9 +614,6 @@ void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
 			ttl = ip4_dst_hoplimit(&rt->dst);
 	}
 
-	if (!df && skb->protocol == htons(ETH_P_IP))
-		df = inner_iph->frag_off & htons(IP_DF);
-
 	headroom += LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len;
 	if (headroom > dev->needed_headroom)
 		dev->needed_headroom = headroom;