diff mbox

[RFC] gretap: don't do MTU discovery

Message ID 20121026160515.1f37592e@nehalam.linuxnetplumber.net
State RFC, archived
Delegated to: David Miller
Headers show

Commit Message

stephen hemminger Oct. 26, 2012, 11:05 p.m. UTC 
This came up during review of VXLAN.

Since gretap is a L2 tunnel, and it is not necessarily in the same IP
network space as there is no way to safely do MTU discovery.
The IP address in the inner header may not be reachable or conflict
with the IP address visible to the kernel.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

--- a/net/ipv4/ip_gre.c	2012-10-26 15:50:39.579955151 -0700
+++ b/net/ipv4/ip_gre.c	2012-10-26 15:58:31.667213979 -0700
@@ -765,7 +765,6 @@  static netdev_tx_t ipgre_tunnel_xmit(str
 	unsigned int max_headroom;		/* The extra header space needed */
 	int    gre_hlen;
 	__be32 dst;
-	int    mtu;
 
 	if (skb->ip_summed == CHECKSUM_PARTIAL &&
 	    skb_checksum_help(skb))
@@ -853,44 +852,54 @@  static netdev_tx_t ipgre_tunnel_xmit(str
 	}
 
 	df = tiph->frag_off;
-	if (df)
-		mtu = dst_mtu(&rt->dst) - dev->hard_header_len - tunnel->hlen;
-	else
-		mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
-
-	if (skb_dst(skb))
-		skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu);
-
-	if (skb->protocol == htons(ETH_P_IP)) {
-		df |= (old_iph->frag_off&htons(IP_DF));
-
-		if ((old_iph->frag_off&htons(IP_DF)) &&
-		    mtu < ntohs(old_iph->tot_len)) {
-			icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
-			ip_rt_put(rt);
-			goto tx_error;
+	if (dev->type == ARPHRD_IPGRE) {
+		unsigned int mtu;
+
+		if (df)
+			mtu = dst_mtu(&rt->dst)
+				- dev->hard_header_len - tunnel->hlen;
+		else
+			mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
+
+		if (skb_dst(skb))
+			skb_dst(skb)->ops->update_pmtu(skb_dst(skb),
+						       NULL, skb, mtu);
+
+		if (skb->protocol == htons(ETH_P_IP)) {
+			df |= (old_iph->frag_off&htons(IP_DF));
+
+			if ((old_iph->frag_off&htons(IP_DF)) &&
+			    mtu < ntohs(old_iph->tot_len)) {
+				icmp_send(skb, ICMP_DEST_UNREACH,
+					  ICMP_FRAG_NEEDED, htonl(mtu));
+				ip_rt_put(rt);
+				goto tx_error;
+			}
 		}
-	}
 #if IS_ENABLED(CONFIG_IPV6)
-	else if (skb->protocol == htons(ETH_P_IPV6)) {
-		struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
+		else if (skb->protocol == htons(ETH_P_IPV6)) {
+			struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
 
-		if (rt6 && mtu < dst_mtu(skb_dst(skb)) && mtu >= IPV6_MIN_MTU) {
-			if ((tunnel->parms.iph.daddr &&
-			     !ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
-			    rt6->rt6i_dst.plen == 128) {
-				rt6->rt6i_flags |= RTF_MODIFIED;
-				dst_metric_set(skb_dst(skb), RTAX_MTU, mtu);
+			if (rt6 && mtu < dst_mtu(skb_dst(skb)) &&
+			    mtu >= IPV6_MIN_MTU) {
+				if ((tunnel->parms.iph.daddr &&
+				     !ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
+				    rt6->rt6i_dst.plen == 128) {
+					rt6->rt6i_flags |= RTF_MODIFIED;
+					dst_metric_set(skb_dst(skb),
+						       RTAX_MTU, mtu);
+				}
 			}
-		}
 
-		if (mtu >= IPV6_MIN_MTU && mtu < skb->len - tunnel->hlen + gre_hlen) {
-			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
-			ip_rt_put(rt);
-			goto tx_error;
+			if (mtu >= IPV6_MIN_MTU &&
+			    mtu < skb->len - tunnel->hlen + gre_hlen) {
+				icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+				ip_rt_put(rt);
+				goto tx_error;
+			}
 		}
-	}
 #endif
+	}
 
 	if (tunnel->err_count > 0) {
 		if (time_before(jiffies,