Skip to contentSkip to conversation

FAILED: patch "[PATCH] net: ip_tunnel: fix mtu calculation for ETHER tunnel devices" failed to apply to 4.19-stable tree

PATCH Applied [email protected] 2 participants 3 messages Jul 22, 2021 – Jul 23, 2021

Subsystems (3)
Networking Stable branch IPv4/IPv6
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 4.19-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 9992a078b1771da354ac1f9737e1e639b687caa2 Mon Sep 17 00:00:00 2001
From: Hangbin Liu <[email protected]>
Date: Fri, 9 Jul 2021 11:45:02 +0800
Subject: [PATCH] net: ip_tunnel: fix mtu calculation for ETHER tunnel devices

Commit 28e104d00281 ("net: ip_tunnel: fix mtu calculation") removed dev->hard_header_len subtraction when calculate MTU for tunnel devices as there is an overhead for device that has header_ops.

But there are ETHER tunnel devices, like gre_tap or erspan, which don't have header_ops but set dev->hard_header_len during setup. This makes pkts greater than (MTU - ETH_HLEN) could not be xmited. Fix it by subtracting the ETHER tunnel devices' dev->hard_header_len for MTU calculation.

Fixes 28e104d00281 ("net: ip_tunnel: fix mtu calculation")
Reported-by Jianlin Shi <[email protected]>
Signed-off-by Hangbin Liu <[email protected]>
Signed-off-by David S. Miller <[email protected]>
net/ipv4/ip_tunnel.c +15 -3
index f6cc26de5ed3..0dca00745ac3 100644
@@ -317,7 +317,7 @@ static int ip_tunnel_bind_dev(struct net_device *dev)
 	}
 
 	dev->needed_headroom = t_hlen + hlen;
-	mtu -= t_hlen;
+	mtu -= t_hlen + (dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0);
 
 	if (mtu < IPV4_MIN_MTU)
 		mtu = IPV4_MIN_MTU;

@@ -348,6 +348,9 @@ static struct ip_tunnel *ip_tunnel_create(struct net *net,
 	t_hlen = nt->hlen + sizeof(struct iphdr);
 	dev->min_mtu = ETH_MIN_MTU;
 	dev->max_mtu = IP_MAX_MTU - t_hlen;
+	if (dev->type == ARPHRD_ETHER)
+		dev->max_mtu -= dev->hard_header_len;
+
 	ip_tunnel_add(itn, nt);
 	return nt;
 

@@ -489,11 +492,14 @@ static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
 
 	tunnel_hlen = md ? tunnel_hlen : tunnel->hlen;
 	pkt_size = skb->len - tunnel_hlen;
+	pkt_size -= dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0;
 
-	if (df)
+	if (df) {
 		mtu = dst_mtu(&rt->dst) - (sizeof(struct iphdr) + tunnel_hlen);
-	else
+		mtu -= dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0;
+	} else {
 		mtu = skb_valid_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
+	}
 
 	if (skb_valid_dst(skb))
 		skb_dst_update_pmtu_no_confirm(skb, mtu);

@@ -972,6 +978,9 @@ int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict)
 	int t_hlen = tunnel->hlen + sizeof(struct iphdr);
 	int max_mtu = IP_MAX_MTU - t_hlen;
 
+	if (dev->type == ARPHRD_ETHER)
+		max_mtu -= dev->hard_header_len;
+
 	if (new_mtu < ETH_MIN_MTU)
 		return -EINVAL;
 

@@ -1149,6 +1158,9 @@ int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[],
 	if (tb[IFLA_MTU]) {
 		unsigned int max = IP_MAX_MTU - (nt->hlen + sizeof(struct iphdr));
 
+		if (dev->type == ARPHRD_ETHER)
+			max -= dev->hard_header_len;
+
 		mtu = clamp(dev->mtu, (unsigned int)ETH_MIN_MTU, max);
 	}
#1 lore raw

PATCH 1/1 net: ip_tunnel: fix mtu calculation for ETHER tunnel devices

commit 9992a078b1771da354ac1f9737e1e639b687caa2 upstream.

Commit 28e104d00281 ("net: ip_tunnel: fix mtu calculation") removed dev->hard_header_len subtraction when calculate MTU for tunnel devices as there is an overhead for device that has header_ops.

But there are ETHER tunnel devices, like gre_tap or erspan, which don't have header_ops but set dev->hard_header_len during setup. This makes pkts greater than (MTU - ETH_HLEN) could not be xmited. Fix it by subtracting the ETHER tunnel devices' dev->hard_header_len for MTU calculation.

Fixes 28e104d00281 ("net: ip_tunnel: fix mtu calculation")
Reported-by Jianlin Shi <[email protected]>
Signed-off-by Hangbin Liu <[email protected]>
Signed-off-by David S. Miller <[email protected]>
---
 net/ipv4/ip_tunnel.c | 22 ++++++++++++++++++----
 1 file changed, 18 insertions(+), 4 deletions(-)
net/ipv4/ip_tunnel.c +18 -4
index bdd073ea300a..30e93b4f831f 100644
@@ -330,7 +330,7 @@ static int ip_tunnel_bind_dev(struct net_device *dev)
 	}
 
 	dev->needed_headroom = t_hlen + hlen;
-	mtu -= t_hlen;
+	mtu -= t_hlen + (dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0);
 
 	if (mtu < IPV4_MIN_MTU)
 		mtu = IPV4_MIN_MTU;

@@ -361,6 +361,9 @@ static struct ip_tunnel *ip_tunnel_create(struct net *net,
 	t_hlen = nt->hlen + sizeof(struct iphdr);
 	dev->min_mtu = ETH_MIN_MTU;
 	dev->max_mtu = IP_MAX_MTU - t_hlen;
+	if (dev->type == ARPHRD_ETHER)
+		dev->max_mtu -= dev->hard_header_len;
+
 	ip_tunnel_add(itn, nt);
 	return nt;
 

@@ -502,13 +505,18 @@ static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
 			    const struct iphdr *inner_iph)
 {
 	struct ip_tunnel *tunnel = netdev_priv(dev);
-	int pkt_size = skb->len - tunnel->hlen;
+	int pkt_size;
 	int mtu;
 
-	if (df)
+	pkt_size = skb->len - tunnel->hlen;
+	pkt_size -= dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0;
+
+	if (df) {
 		mtu = dst_mtu(&rt->dst) - (sizeof(struct iphdr) + tunnel->hlen);
-	else
+		mtu -= dev->type == ARPHRD_ETHER ? dev->hard_header_len : 0;
+	} else {
 		mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
+	}
 
 	skb_dst_update_pmtu_no_confirm(skb, mtu);
 

@@ -936,6 +944,9 @@ int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict)
 	int t_hlen = tunnel->hlen + sizeof(struct iphdr);
 	int max_mtu = IP_MAX_MTU - t_hlen;
 
+	if (dev->type == ARPHRD_ETHER)
+		max_mtu -= dev->hard_header_len;
+
 	if (new_mtu < ETH_MIN_MTU)
 		return -EINVAL;
 

@@ -1113,6 +1124,9 @@ int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[],
 	if (tb[IFLA_MTU]) {
 		unsigned int max = IP_MAX_MTU - (nt->hlen + sizeof(struct iphdr));
 
+		if (dev->type == ARPHRD_ETHER)
+			max -= dev->hard_header_len;
+
 		mtu = clamp(dev->mtu, (unsigned int)ETH_MIN_MTU, max);
 	}
Signature · 1 line
2.18.1
author maintainer Applied
#2 lore raw

Re: [PATCH 4.19] net: ip_tunnel: fix mtu calculation for ETHER tunnel devices

Quoted 12 lines · Hangbin Liu

commit 9992a078b1771da354ac1f9737e1e639b687caa2 upstream.

Commit 28e104d00281 ("net: ip_tunnel: fix mtu calculation") removed dev->hard_header_len subtraction when calculate MTU for tunnel devices as there is an overhead for device that has header_ops.

But there are ETHER tunnel devices, like gre_tap or erspan, which don't have header_ops but set dev->hard_header_len during setup. This makes pkts greater than (MTU - ETH_HLEN) could not be xmited. Fix it by subtracting the ETHER tunnel devices' dev->hard_header_len for MTU calculation.

Quoted 7 lines
Fixes 28e104d00281 ("net: ip_tunnel: fix mtu calculation")
Reported-by Jianlin Shi <[email protected]>
Signed-off-by Hangbin Liu <[email protected]>
Signed-off-by David S. Miller <[email protected]>
---
 net/ipv4/ip_tunnel.c | 22 ++++++++++++++++++----
 1 file changed, 18 insertions(+), 4 deletions(-)

Applied, thanks.

greg k-h