[PATCH net,v2] ip6_tunnel, ip6_gre: fix setting of DSCP on encapsulated packets

[Peter Dawson]

This fix addresses two problems in the way the DSCP field is formulated
 on the encapsulating header of IPv6 tunnels.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=195661

1) The IPv6 tunneling code was manipulating the DSCP field of the
 encapsulating packet using the 32b flowlabel. Since the flowlabel is
 only the lower 20b it was incorrect to assume that the upper 12b
 containing the DSCP and ECN fields would remain intact when formulating
 the encapsulating header. This fix handles the 'inherit' and
 'fixed-value' DSCP cases explicitly using the extant dsfield u8 variable.

2) The use of INET_ECN_encapsulate(0, dsfield) in ip6_tnl_xmit was
 incorrect and resulted in the DSCP value always being set to 0.

Commit 90427ef5d2a4 ("ipv6: fix flow labels when the traffic class 
 is non-0") caused the regression by masking out the flowlabel
 which exposed the incorrect handling of the DSCP portion of the 
 flowlabel in ip6_tunnel and ip6_gre.

Fixes: 90427ef5d2a4 ("ipv6: fix flow labels when the traffic class is non-0")
Signed-off-by: Peter Dawson <peter.a.dawson@xxxxxxxxxx>
---
 net/ipv6/ip6_gre.c    | 13 +++++++------
 net/ipv6/ip6_tunnel.c | 21 +++++++++++++--------
 2 files changed, 20 insertions(+), 14 deletions(-)

diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
index 8d128ba..0c5b4caa 100644
--- a/net/ipv6/ip6_gre.c
+++ b/net/ipv6/ip6_gre.c
@@ -537,11 +537,10 @@ static inline int ip6gre_xmit_ipv4(struct sk_buff *skb, struct net_device *dev)
 
 	memcpy(&fl6, &t->fl.u.ip6, sizeof(fl6));
 
-	dsfield = ipv4_get_dsfield(iph);
-
 	if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)
-		fl6.flowlabel |= htonl((__u32)iph->tos << IPV6_TCLASS_SHIFT)
-					  & IPV6_TCLASS_MASK;
+		dsfield = ipv4_get_dsfield(iph);
+	else
+		dsfield = ip6_tclass(t->parms.flowinfo);
 	if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK)
 		fl6.flowi6_mark = skb->mark;
 	else
@@ -598,9 +597,11 @@ static inline int ip6gre_xmit_ipv6(struct sk_buff *skb, struct net_device *dev)
 
 	memcpy(&fl6, &t->fl.u.ip6, sizeof(fl6));
 
-	dsfield = ipv6_get_dsfield(ipv6h);
 	if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)
-		fl6.flowlabel |= (*(__be32 *) ipv6h & IPV6_TCLASS_MASK);
+		dsfield = ipv6_get_dsfield(ipv6h);
+	else
+		dsfield = ip6_tclass(t->parms.flowinfo);
+
 	if (t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL)
 		fl6.flowlabel |= ip6_flowlabel(ipv6h);
 	if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK)
diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c
index 6eb2ae5..7ae6c50 100644
--- a/net/ipv6/ip6_tunnel.c
+++ b/net/ipv6/ip6_tunnel.c
@@ -1196,7 +1196,7 @@ int ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev, __u8 dsfield,
 	skb_push(skb, sizeof(struct ipv6hdr));
 	skb_reset_network_header(skb);
 	ipv6h = ipv6_hdr(skb);
-	ip6_flow_hdr(ipv6h, INET_ECN_encapsulate(0, dsfield),
+	ip6_flow_hdr(ipv6h, dsfield,
 		     ip6_make_flowlabel(net, skb, fl6->flowlabel, true, fl6));
 	ipv6h->hop_limit = hop_limit;
 	ipv6h->nexthdr = proto;
@@ -1231,8 +1231,6 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 	if (tproto != IPPROTO_IPIP && tproto != 0)
 		return -1;
 
-	dsfield = ipv4_get_dsfield(iph);
-
 	if (t->parms.collect_md) {
 		struct ip_tunnel_info *tun_info;
 		const struct ip_tunnel_key *key;
@@ -1246,6 +1244,7 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 		fl6.flowi6_proto = IPPROTO_IPIP;
 		fl6.daddr = key->u.ipv6.dst;
 		fl6.flowlabel = key->label;
+		dsfield = ip6_tclass(key->label);
 	} else {
 		if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT))
 			encap_limit = t->parms.encap_limit;
@@ -1254,8 +1253,9 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 		fl6.flowi6_proto = IPPROTO_IPIP;
 
 		if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)
-			fl6.flowlabel |= htonl((__u32)iph->tos << IPV6_TCLASS_SHIFT)
-					 & IPV6_TCLASS_MASK;
+			dsfield = ipv4_get_dsfield(iph);
+		else
+			dsfield = ip6_tclass(t->parms.flowinfo);
 		if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK)
 			fl6.flowi6_mark = skb->mark;
 		else
@@ -1267,6 +1267,8 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 	if (iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6))
 		return -1;
 
+	dsfield = INET_ECN_encapsulate(dsfield, ipv4_get_dsfield(iph));
+
 	skb_set_inner_ipproto(skb, IPPROTO_IPIP);
 
 	err = ip6_tnl_xmit(skb, dev, dsfield, &fl6, encap_limit, &mtu,
@@ -1300,8 +1302,6 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 	    ip6_tnl_addr_conflict(t, ipv6h))
 		return -1;
 
-	dsfield = ipv6_get_dsfield(ipv6h);
-
 	if (t->parms.collect_md) {
 		struct ip_tunnel_info *tun_info;
 		const struct ip_tunnel_key *key;
@@ -1315,6 +1315,7 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 		fl6.flowi6_proto = IPPROTO_IPV6;
 		fl6.daddr = key->u.ipv6.dst;
 		fl6.flowlabel = key->label;
+		dsfield = ip6_tclass(key->label);
 	} else {
 		offset = ip6_tnl_parse_tlv_enc_lim(skb, skb_network_header(skb));
 		/* ip6_tnl_parse_tlv_enc_lim() might have reallocated skb->head */
@@ -1337,7 +1338,9 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 		fl6.flowi6_proto = IPPROTO_IPV6;
 
 		if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)
-			fl6.flowlabel |= (*(__be32 *)ipv6h & IPV6_TCLASS_MASK);
+			dsfield = ipv6_get_dsfield(ipv6h);
+		else
+			dsfield = ip6_tclass(t->parms.flowinfo);
 		if (t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL)
 			fl6.flowlabel |= ip6_flowlabel(ipv6h);
 		if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK)
@@ -1351,6 +1354,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
 	if (iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6))
 		return -1;
 
+	dsfield = INET_ECN_encapsulate(dsfield, ipv6_get_dsfield(ipv6h));
+
 	skb_set_inner_ipproto(skb, IPPROTO_IPV6);
 
 	err = ip6_tnl_xmit(skb, dev, dsfield, &fl6, encap_limit, &mtu,
-- 
2.7.4