kernel: port upstream nft_flow_offload changes to xt_FLOWOFFLOAD and fix routing issues

Replace an old cleanup patch that never made it upstream with the proper
upstream fix. This patch was incompatible with the recent changes that
affected the way that the flow tuple dst entry was used.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commits 442ecce76169d and c8933ce533656)

4 files changed, 127 insertions, 91 deletions

diff --git a/target/linux/generic/backport-4.14/366-netfilter-nf_flow_table-clean-up-and-fix-dst-handlin.patch b/target/linux/generic/backport-4.14/366-netfilter-nf_flow_table-clean-up-and-fix-dst-handlin.patch
deleted file mode 100644
index 491f057858..0000000000
--- a/target/linux/generic/backport-4.14/366-netfilter-nf_flow_table-clean-up-and-fix-dst-handlin.patch
+++ /dev/null
@@ -1,89 +0,0 @@
-From: Felix Fietkau <nbd@nbd.name>
-Date: Thu, 15 Mar 2018 18:21:43 +0100
-Subject: [PATCH] netfilter: nf_flow_table: clean up and fix dst handling
-
-dst handling in the code is inconsistent and possibly wrong. In my test,
-skb_dst(skb) holds the dst entry after routing but before NAT, so the
-code could possibly return the same dst entry for both directions of a
-connection.
-Additionally, there was some confusion over the dst entry vs the address
-passed as parameter to rt_nexthop/rt6_nexthop.
-
-Do an explicit dst lookup for both ends of the connection and always use
-the source address for it. When running the IP hook, use the dst entry
-for the opposite direction for determining the route.
-
-Signed-off-by: Felix Fietkau <nbd@nbd.name>
----
-
---- a/net/netfilter/nf_flow_table_ip.c
-+++ b/net/netfilter/nf_flow_table_ip.c
-@@ -238,7 +238,7 @@ nf_flow_offload_ip_hook(void *priv, stru
- 
- 	dir = tuplehash->tuple.dir;
- 	flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]);
--	rt = (const struct rtable *)flow->tuplehash[dir].tuple.dst_cache;
-+	rt = (const struct rtable *)flow->tuplehash[!dir].tuple.dst_cache;
- 
- 	if (unlikely(nf_flow_exceeds_mtu(skb, flow->tuplehash[dir].tuple.mtu)) &&
- 	    (ip_hdr(skb)->frag_off & htons(IP_DF)) != 0)
-@@ -455,7 +455,7 @@ nf_flow_offload_ipv6_hook(void *priv, st
- 
- 	dir = tuplehash->tuple.dir;
- 	flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]);
--	rt = (struct rt6_info *)flow->tuplehash[dir].tuple.dst_cache;
-+	rt = (struct rt6_info *)flow->tuplehash[!dir].tuple.dst_cache;
- 
- 	if (unlikely(nf_flow_exceeds_mtu(skb, flow->tuplehash[dir].tuple.mtu)))
- 		return NF_ACCEPT;
---- a/net/netfilter/nft_flow_offload.c
-+++ b/net/netfilter/nft_flow_offload.c
-@@ -17,27 +17,38 @@ struct nft_flow_offload {
- 	struct nft_flowtable	*flowtable;
- };
- 
--static int nft_flow_route(const struct nft_pktinfo *pkt,
--			  const struct nf_conn *ct,
--			  struct nf_flow_route *route,
--			  enum ip_conntrack_dir dir)
-+static struct dst_entry *
-+nft_flow_dst(const struct nf_conn *ct, enum ip_conntrack_dir dir,
-+	     const struct nft_pktinfo *pkt)
- {
--	struct dst_entry *this_dst = skb_dst(pkt->skb);
--	struct dst_entry *other_dst = NULL;
-+	struct dst_entry *dst;
- 	struct flowi fl;
- 
- 	memset(&fl, 0, sizeof(fl));
- 	switch (nft_pf(pkt)) {
- 	case NFPROTO_IPV4:
--		fl.u.ip4.daddr = ct->tuplehash[!dir].tuple.dst.u3.ip;
-+		fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
- 		break;
- 	case NFPROTO_IPV6:
--		fl.u.ip6.daddr = ct->tuplehash[!dir].tuple.dst.u3.in6;
-+		fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
- 		break;
- 	}
- 
--	nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt));
--	if (!other_dst)
-+	nf_route(nft_net(pkt), &dst, &fl, false, nft_pf(pkt));
-+
-+	return dst;
-+}
-+
-+static int nft_flow_route(const struct nft_pktinfo *pkt,
-+			  const struct nf_conn *ct,
-+			  struct nf_flow_route *route,
-+			  enum ip_conntrack_dir dir)
-+{
-+	struct dst_entry *this_dst, *other_dst;
-+
-+	this_dst = nft_flow_dst(ct, dir, pkt);
-+	other_dst = nft_flow_dst(ct, !dir, pkt);
-+	if (!this_dst || !other_dst)
- 		return -ENOENT;
- 
- 	route->tuple[dir].dst		= this_dst;
diff --git a/target/linux/generic/backport-4.14/366-netfilter-nft_flow_offload-Fix-reverse-route-lookup.patch b/target/linux/generic/backport-4.14/366-netfilter-nft_flow_offload-Fix-reverse-route-lookup.patch
new file mode 100644
index 0000000000..e91aaa91d7
--- /dev/null
+++ b/target/linux/generic/backport-4.14/366-netfilter-nft_flow_offload-Fix-reverse-route-lookup.patch
@@ -0,0 +1,39 @@
+From: wenxu <wenxu@ucloud.cn>
+Date: Wed, 9 Jan 2019 10:40:11 +0800
+Subject: [PATCH] netfilter: nft_flow_offload: Fix reverse route lookup
+
+Using the following example:
+
+	client 1.1.1.7 ---> 2.2.2.7 which dnat to 10.0.0.7 server
+
+The first reply packet (ie. syn+ack) uses an incorrect destination
+address for the reverse route lookup since it uses:
+
+	daddr = ct->tuplehash[!dir].tuple.dst.u3.ip;
+
+which is 2.2.2.7 in the scenario that is described above, while this
+should be:
+
+	daddr = ct->tuplehash[dir].tuple.src.u3.ip;
+
+that is 10.0.0.7.
+
+Signed-off-by: wenxu <wenxu@ucloud.cn>
+Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
+---
+
+--- a/net/netfilter/nft_flow_offload.c
++++ b/net/netfilter/nft_flow_offload.c
+@@ -29,10 +29,10 @@ static int nft_flow_route(const struct n
+ 	memset(&fl, 0, sizeof(fl));
+ 	switch (nft_pf(pkt)) {
+ 	case NFPROTO_IPV4:
+-		fl.u.ip4.daddr = ct->tuplehash[!dir].tuple.dst.u3.ip;
++		fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
+ 		break;
+ 	case NFPROTO_IPV6:
+-		fl.u.ip6.daddr = ct->tuplehash[!dir].tuple.dst.u3.in6;
++		fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
+ 		break;
+ 	}
+ 
diff --git a/target/linux/generic/backport-4.14/369-v4.18-netfilter-nf_flow_table-attach-dst-to-skbs.patch b/target/linux/generic/backport-4.14/369-v4.18-netfilter-nf_flow_table-attach-dst-to-skbs.patch
index 35d099097a..ad72d65c77 100644
--- a/target/linux/generic/backport-4.14/369-v4.18-netfilter-nf_flow_table-attach-dst-to-skbs.patch
+++ b/target/linux/generic/backport-4.14/369-v4.18-netfilter-nf_flow_table-attach-dst-to-skbs.patch
@@ -26,8 +26,8 @@ Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
  
  	dir = tuplehash->tuple.dir;
  	flow = container_of(tuplehash, struct flow_offload, tuplehash[dir]);
--	rt = (const struct rtable *)flow->tuplehash[!dir].tuple.dst_cache;
-+	rt = (struct rtable *)flow->tuplehash[!dir].tuple.dst_cache;
+-	rt = (const struct rtable *)flow->tuplehash[dir].tuple.dst_cache;
++	rt = (struct rtable *)flow->tuplehash[dir].tuple.dst_cache;
  
  	if (unlikely(nf_flow_exceeds_mtu(skb, flow->tuplehash[dir].tuple.mtu)) &&
  	    (ip_hdr(skb)->frag_off & htons(IP_DF)) != 0)
diff --git a/target/linux/generic/backport-4.14/372-netfilter-nft_flow_offload-fix-interaction-with-vrf-.patch b/target/linux/generic/backport-4.14/372-netfilter-nft_flow_offload-fix-interaction-with-vrf-.patch
new file mode 100644
index 0000000000..f0a8b19eec
--- /dev/null
+++ b/target/linux/generic/backport-4.14/372-netfilter-nft_flow_offload-fix-interaction-with-vrf-.patch
@@ -0,0 +1,86 @@
+From: wenxu <wenxu@ucloud.cn>
+Date: Thu, 10 Jan 2019 14:51:35 +0800
+Subject: [PATCH] netfilter: nft_flow_offload: fix interaction with vrf slave
+ device
+
+In the forward chain, the iif is changed from slave device to master vrf
+device. Thus, flow offload does not find a match on the lower slave
+device.
+
+This patch uses the cached route, ie. dst->dev, to update the iif and
+oif fields in the flow entry.
+
+After this patch, the following example works fine:
+
+ # ip addr add dev eth0 1.1.1.1/24
+ # ip addr add dev eth1 10.0.0.1/24
+ # ip link add user1 type vrf table 1
+ # ip l set user1 up
+ # ip l set dev eth0 master user1
+ # ip l set dev eth1 master user1
+
+ # nft add table firewall
+ # nft add flowtable f fb1 { hook ingress priority 0 \; devices = { eth0, eth1 } \; }
+ # nft add chain f ftb-all {type filter hook forward priority 0 \; policy accept \; }
+ # nft add rule f ftb-all ct zone 1 ip protocol tcp flow offload @fb1
+ # nft add rule f ftb-all ct zone 1 ip protocol udp flow offload @fb1
+
+Signed-off-by: wenxu <wenxu@ucloud.cn>
+Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
+---
+
+--- a/include/net/netfilter/nf_flow_table.h
++++ b/include/net/netfilter/nf_flow_table.h
+@@ -84,7 +84,6 @@ struct flow_offload {
+ struct nf_flow_route {
+ 	struct {
+ 		struct dst_entry	*dst;
+-		int			ifindex;
+ 	} tuple[FLOW_OFFLOAD_DIR_MAX];
+ };
+ 
+--- a/net/netfilter/nf_flow_table_core.c
++++ b/net/netfilter/nf_flow_table_core.c
+@@ -28,6 +28,7 @@ flow_offload_fill_dir(struct flow_offloa
+ {
+ 	struct flow_offload_tuple *ft = &flow->tuplehash[dir].tuple;
+ 	struct nf_conntrack_tuple *ctt = &ct->tuplehash[dir].tuple;
++	struct dst_entry *other_dst = route->tuple[!dir].dst;
+ 	struct dst_entry *dst = route->tuple[dir].dst;
+ 
+ 	ft->dir = dir;
+@@ -50,8 +51,8 @@ flow_offload_fill_dir(struct flow_offloa
+ 	ft->src_port = ctt->src.u.tcp.port;
+ 	ft->dst_port = ctt->dst.u.tcp.port;
+ 
+-	ft->iifidx = route->tuple[dir].ifindex;
+-	ft->oifidx = route->tuple[!dir].ifindex;
++	ft->iifidx = other_dst->dev->ifindex;
++	ft->oifidx = dst->dev->ifindex;
+ 	ft->dst_cache = dst;
+ }
+ 
+--- a/net/netfilter/nft_flow_offload.c
++++ b/net/netfilter/nft_flow_offload.c
+@@ -30,9 +30,11 @@ static int nft_flow_route(const struct n
+ 	switch (nft_pf(pkt)) {
+ 	case NFPROTO_IPV4:
+ 		fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
++		fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex;
+ 		break;
+ 	case NFPROTO_IPV6:
+ 		fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
++		fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex;
+ 		break;
+ 	}
+ 
+@@ -41,9 +43,7 @@ static int nft_flow_route(const struct n
+ 		return -ENOENT;
+ 
+ 	route->tuple[dir].dst		= this_dst;
+-	route->tuple[dir].ifindex	= nft_in(pkt)->ifindex;
+ 	route->tuple[!dir].dst		= other_dst;
+-	route->tuple[!dir].ifindex	= nft_out(pkt)->ifindex;
+ 
+ 	return 0;
+ }