1 files changed, 222 insertions, 49 deletions

diff --git a/net/ipv6/seg6_iptunnel.c b/net/ipv6/seg6_iptunnel.c
index d64855010948..3e1b9991131a 100644
--- a/net/ipv6/seg6_iptunnel.c
+++ b/net/ipv6/seg6_iptunnel.c
@@ -36,9 +36,11 @@ static size_t seg6_lwt_headroom(struct seg6_iptunnel_encap *tuninfo)
 	case SEG6_IPTUN_MODE_INLINE:
 		break;
 	case SEG6_IPTUN_MODE_ENCAP:
+	case SEG6_IPTUN_MODE_ENCAP_RED:
 		head = sizeof(struct ipv6hdr);
 		break;
 	case SEG6_IPTUN_MODE_L2ENCAP:
+	case SEG6_IPTUN_MODE_L2ENCAP_RED:
 		return 0;
 	}
 
@@ -122,11 +124,12 @@ static __be32 seg6_make_flowlabel(struct net *net, struct sk_buff *skb,
 	return flowlabel;
 }
 
-/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
-int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
+static int __seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
+			       int proto, struct dst_entry *cache_dst)
 {
 	struct dst_entry *dst = skb_dst(skb);
-	struct net *net = dev_net(dst->dev);
+	struct net_device *dev = dst_dev(dst);
+	struct net *net = dev_net(dev);
 	struct ipv6hdr *hdr, *inner_hdr;
 	struct ipv6_sr_hdr *isrh;
 	int hdrlen, tot_len, err;
@@ -135,7 +138,7 @@ int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
 	hdrlen = (osrh->hdrlen + 1) << 3;
 	tot_len = hdrlen + sizeof(*hdr);
 
-	err = skb_cow_head(skb, tot_len + skb->mac_len);
+	err = skb_cow_head(skb, tot_len + dst_dev_overhead(cache_dst, skb));
 	if (unlikely(err))
 		return err;
 
@@ -179,7 +182,7 @@ int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
 	isrh->nexthdr = proto;
 
 	hdr->daddr = isrh->segments[isrh->first_segment];
-	set_tun_src(net, dst->dev, &hdr->daddr, &hdr->saddr);
+	set_tun_src(net, dev, &hdr->daddr, &hdr->saddr);
 
 #ifdef CONFIG_IPV6_SEG6_HMAC
 	if (sr_has_hmac(isrh)) {
@@ -189,14 +192,142 @@ int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
 	}
 #endif
 
+	hdr->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
+
 	skb_postpush_rcsum(skb, hdr, tot_len);
 
 	return 0;
 }
+
+/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
+int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
+{
+	return __seg6_do_srh_encap(skb, osrh, proto, NULL);
+}
 EXPORT_SYMBOL_GPL(seg6_do_srh_encap);
 
-/* insert an SRH within an IPv6 packet, just after the IPv6 header */
-int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
+/* encapsulate an IPv6 packet within an outer IPv6 header with reduced SRH */
+static int seg6_do_srh_encap_red(struct sk_buff *skb,
+				 struct ipv6_sr_hdr *osrh, int proto,
+				 struct dst_entry *cache_dst)
+{
+	__u8 first_seg = osrh->first_segment;
+	struct dst_entry *dst = skb_dst(skb);
+	struct net_device *dev = dst_dev(dst);
+	struct net *net = dev_net(dev);
+	struct ipv6hdr *hdr, *inner_hdr;
+	int hdrlen = ipv6_optlen(osrh);
+	int red_tlv_offset, tlv_offset;
+	struct ipv6_sr_hdr *isrh;
+	bool skip_srh = false;
+	__be32 flowlabel;
+	int tot_len, err;
+	int red_hdrlen;
+	int tlvs_len;
+
+	if (first_seg > 0) {
+		red_hdrlen = hdrlen - sizeof(struct in6_addr);
+	} else {
+		/* NOTE: if tag/flags and/or other TLVs are introduced in the
+		 * seg6_iptunnel infrastructure, they should be considered when
+		 * deciding to skip the SRH.
+		 */
+		skip_srh = !sr_has_hmac(osrh);
+
+		red_hdrlen = skip_srh ? 0 : hdrlen;
+	}
+
+	tot_len = red_hdrlen + sizeof(struct ipv6hdr);
+
+	err = skb_cow_head(skb, tot_len + dst_dev_overhead(cache_dst, skb));
+	if (unlikely(err))
+		return err;
+
+	inner_hdr = ipv6_hdr(skb);
+	flowlabel = seg6_make_flowlabel(net, skb, inner_hdr);
+
+	skb_push(skb, tot_len);
+	skb_reset_network_header(skb);
+	skb_mac_header_rebuild(skb);
+	hdr = ipv6_hdr(skb);
+
+	/* based on seg6_do_srh_encap() */
+	if (skb->protocol == htons(ETH_P_IPV6)) {
+		ip6_flow_hdr(hdr, ip6_tclass(ip6_flowinfo(inner_hdr)),
+			     flowlabel);
+		hdr->hop_limit = inner_hdr->hop_limit;
+	} else {
+		ip6_flow_hdr(hdr, 0, flowlabel);
+		hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
+
+		memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
+		IP6CB(skb)->iif = skb->skb_iif;
+	}
+
+	/* no matter if we have to skip the SRH or not, the first segment
+	 * always comes in the pushed IPv6 header.
+	 */
+	hdr->daddr = osrh->segments[first_seg];
+
+	if (skip_srh) {
+		hdr->nexthdr = proto;
+
+		set_tun_src(net, dev, &hdr->daddr, &hdr->saddr);
+		goto out;
+	}
+
+	/* we cannot skip the SRH, slow path */
+
+	hdr->nexthdr = NEXTHDR_ROUTING;
+	isrh = (void *)hdr + sizeof(struct ipv6hdr);
+
+	if (unlikely(!first_seg)) {
+		/* this is a very rare case; we have only one SID but
+		 * we cannot skip the SRH since we are carrying some
+		 * other info.
+		 */
+		memcpy(isrh, osrh, hdrlen);
+		goto srcaddr;
+	}
+
+	tlv_offset = sizeof(*osrh) + (first_seg + 1) * sizeof(struct in6_addr);
+	red_tlv_offset = tlv_offset - sizeof(struct in6_addr);
+
+	memcpy(isrh, osrh, red_tlv_offset);
+
+	tlvs_len = hdrlen - tlv_offset;
+	if (unlikely(tlvs_len > 0)) {
+		const void *s = (const void *)osrh + tlv_offset;
+		void *d = (void *)isrh + red_tlv_offset;
+
+		memcpy(d, s, tlvs_len);
+	}
+
+	--isrh->first_segment;
+	isrh->hdrlen -= 2;
+
+srcaddr:
+	isrh->nexthdr = proto;
+	set_tun_src(net, dev, &hdr->daddr, &hdr->saddr);
+
+#ifdef CONFIG_IPV6_SEG6_HMAC
+	if (unlikely(!skip_srh && sr_has_hmac(isrh))) {
+		err = seg6_push_hmac(net, &hdr->saddr, isrh);
+		if (unlikely(err))
+			return err;
+	}
+#endif
+
+out:
+	hdr->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
+
+	skb_postpush_rcsum(skb, hdr, tot_len);
+
+	return 0;
+}
+
+static int __seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
+				struct dst_entry *cache_dst)
 {
 	struct ipv6hdr *hdr, *oldhdr;
 	struct ipv6_sr_hdr *isrh;
@@ -204,7 +335,7 @@ int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
 
 	hdrlen = (osrh->hdrlen + 1) << 3;
 
-	err = skb_cow_head(skb, hdrlen + skb->mac_len);
+	err = skb_cow_head(skb, hdrlen + dst_dev_overhead(cache_dst, skb));
 	if (unlikely(err))
 		return err;
 
@@ -233,7 +364,7 @@ int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
 
 #ifdef CONFIG_IPV6_SEG6_HMAC
 	if (sr_has_hmac(isrh)) {
-		struct net *net = dev_net(skb_dst(skb)->dev);
+		struct net *net = skb_dst_dev_net(skb);
 
 		err = seg6_push_hmac(net, &hdr->saddr, isrh);
 		if (unlikely(err))
@@ -241,13 +372,14 @@ int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
 	}
 #endif
 
+	hdr->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
+
 	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);
 
 	return 0;
 }
-EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
 
-static int seg6_do_srh(struct sk_buff *skb)
+static int seg6_do_srh(struct sk_buff *skb, struct dst_entry *cache_dst)
 {
 	struct dst_entry *dst = skb_dst(skb);
 	struct seg6_iptunnel_encap *tinfo;
@@ -260,11 +392,12 @@ static int seg6_do_srh(struct sk_buff *skb)
 		if (skb->protocol != htons(ETH_P_IPV6))
 			return -EINVAL;
 
-		err = seg6_do_srh_inline(skb, tinfo->srh);
+		err = __seg6_do_srh_inline(skb, tinfo->srh, cache_dst);
 		if (err)
 			return err;
 		break;
 	case SEG6_IPTUN_MODE_ENCAP:
+	case SEG6_IPTUN_MODE_ENCAP_RED:
 		err = iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6);
 		if (err)
 			return err;
@@ -276,7 +409,13 @@ static int seg6_do_srh(struct sk_buff *skb)
 		else
 			return -EINVAL;
 
-		err = seg6_do_srh_encap(skb, tinfo->srh, proto);
+		if (tinfo->mode == SEG6_IPTUN_MODE_ENCAP)
+			err = __seg6_do_srh_encap(skb, tinfo->srh,
+						  proto, cache_dst);
+		else
+			err = seg6_do_srh_encap_red(skb, tinfo->srh,
+						    proto, cache_dst);
+
 		if (err)
 			return err;
 
@@ -285,6 +424,7 @@ static int seg6_do_srh(struct sk_buff *skb)
 		skb->protocol = htons(ETH_P_IPV6);
 		break;
 	case SEG6_IPTUN_MODE_L2ENCAP:
+	case SEG6_IPTUN_MODE_L2ENCAP_RED:
 		if (!skb_mac_header_was_set(skb))
 			return -EINVAL;
 
@@ -294,7 +434,15 @@ static int seg6_do_srh(struct sk_buff *skb)
 		skb_mac_header_rebuild(skb);
 		skb_push(skb, skb->mac_len);
 
-		err = seg6_do_srh_encap(skb, tinfo->srh, IPPROTO_ETHERNET);
+		if (tinfo->mode == SEG6_IPTUN_MODE_L2ENCAP)
+			err = __seg6_do_srh_encap(skb, tinfo->srh,
+						  IPPROTO_ETHERNET,
+						  cache_dst);
+		else
+			err = seg6_do_srh_encap_red(skb, tinfo->srh,
+						    IPPROTO_ETHERNET,
+						    cache_dst);
+
 		if (err)
 			return err;
 
@@ -302,13 +450,19 @@ static int seg6_do_srh(struct sk_buff *skb)
 		break;
 	}
 
-	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
 	skb_set_transport_header(skb, sizeof(struct ipv6hdr));
 	nf_reset_ct(skb);
 
 	return 0;
 }
 
+/* insert an SRH within an IPv6 packet, just after the IPv6 header */
+int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
+{
+	return __seg6_do_srh_inline(skb, osrh, NULL);
+}
+EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
+
 static int seg6_input_finish(struct net *net, struct sock *sk,
 			     struct sk_buff *skb)
 {
@@ -320,51 +474,63 @@ static int seg6_input_core(struct net *net, struct sock *sk,
 {
 	struct dst_entry *orig_dst = skb_dst(skb);
 	struct dst_entry *dst = NULL;
+	struct lwtunnel_state *lwtst;
 	struct seg6_lwt *slwt;
 	int err;
 
-	err = seg6_do_srh(skb);
-	if (unlikely(err)) {
-		kfree_skb(skb);
-		return err;
-	}
+	/* We cannot dereference "orig_dst" once ip6_route_input() or
+	 * skb_dst_drop() is called. However, in order to detect a dst loop, we
+	 * need the address of its lwtstate. So, save the address of lwtstate
+	 * now and use it later as a comparison.
+	 */
+	lwtst = orig_dst->lwtstate;
 
-	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
+	slwt = seg6_lwt_lwtunnel(lwtst);
 
-	preempt_disable();
+	local_bh_disable();
 	dst = dst_cache_get(&slwt->cache);
-	preempt_enable();
+	local_bh_enable();
 
-	skb_dst_drop(skb);
+	err = seg6_do_srh(skb, dst);
+	if (unlikely(err)) {
+		dst_release(dst);
+		goto drop;
+	}
 
 	if (!dst) {
 		ip6_route_input(skb);
 		dst = skb_dst(skb);
-		if (!dst->error) {
-			preempt_disable();
+
+		/* cache only if we don't create a dst reference loop */
+		if (!dst->error && lwtst != dst->lwtstate) {
+			local_bh_disable();
 			dst_cache_set_ip6(&slwt->cache, dst,
 					  &ipv6_hdr(skb)->saddr);
-			preempt_enable();
+			local_bh_enable();
 		}
+
+		err = skb_cow_head(skb, LL_RESERVED_SPACE(dst_dev(dst)));
+		if (unlikely(err))
+			goto drop;
 	} else {
+		skb_dst_drop(skb);
 		skb_dst_set(skb, dst);
 	}
 
-	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
-	if (unlikely(err))
-		return err;
-
 	if (static_branch_unlikely(&nf_hooks_lwtunnel_enabled))
 		return NF_HOOK(NFPROTO_IPV6, NF_INET_LOCAL_OUT,
 			       dev_net(skb->dev), NULL, skb, NULL,
-			       skb_dst(skb)->dev, seg6_input_finish);
+			       skb_dst_dev(skb), seg6_input_finish);
 
 	return seg6_input_finish(dev_net(skb->dev), NULL, skb);
+drop:
+	kfree_skb(skb);
+	return err;
 }
 
 static int seg6_input_nf(struct sk_buff *skb)
 {
-	struct net_device *dev = skb_dst(skb)->dev;
+	struct net_device *dev = skb_dst_dev(skb);
 	struct net *net = dev_net(skb->dev);
 
 	switch (skb->protocol) {
@@ -395,15 +561,15 @@ static int seg6_output_core(struct net *net, struct sock *sk,
 	struct seg6_lwt *slwt;
 	int err;
 
-	err = seg6_do_srh(skb);
-	if (unlikely(err))
-		goto drop;
-
 	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
 
-	preempt_disable();
+	local_bh_disable();
 	dst = dst_cache_get(&slwt->cache);
-	preempt_enable();
+	local_bh_enable();
+
+	err = seg6_do_srh(skb, dst);
+	if (unlikely(err))
+		goto drop;
 
 	if (unlikely(!dst)) {
 		struct ipv6hdr *hdr = ipv6_hdr(skb);
@@ -419,35 +585,38 @@ static int seg6_output_core(struct net *net, struct sock *sk,
 		dst = ip6_route_output(net, NULL, &fl6);
 		if (dst->error) {
 			err = dst->error;
-			dst_release(dst);
 			goto drop;
 		}
 
-		preempt_disable();
-		dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
-		preempt_enable();
+		/* cache only if we don't create a dst reference loop */
+		if (orig_dst->lwtstate != dst->lwtstate) {
+			local_bh_disable();
+			dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
+			local_bh_enable();
+		}
+
+		err = skb_cow_head(skb, LL_RESERVED_SPACE(dst_dev(dst)));
+		if (unlikely(err))
+			goto drop;
 	}
 
 	skb_dst_drop(skb);
 	skb_dst_set(skb, dst);
 
-	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
-	if (unlikely(err))
-		goto drop;
-
 	if (static_branch_unlikely(&nf_hooks_lwtunnel_enabled))
 		return NF_HOOK(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net, sk, skb,
-			       NULL, skb_dst(skb)->dev, dst_output);
+			       NULL, dst_dev(dst), dst_output);
 
 	return dst_output(net, sk, skb);
 drop:
+	dst_release(dst);
 	kfree_skb(skb);
 	return err;
 }
 
 static int seg6_output_nf(struct net *net, struct sock *sk, struct sk_buff *skb)
 {
-	struct net_device *dev = skb_dst(skb)->dev;
+	struct net_device *dev = skb_dst_dev(skb);
 
 	switch (skb->protocol) {
 	case htons(ETH_P_IP):
@@ -514,6 +683,10 @@ static int seg6_build_state(struct net *net, struct nlattr *nla,
 		break;
 	case SEG6_IPTUN_MODE_L2ENCAP:
 		break;
+	case SEG6_IPTUN_MODE_ENCAP_RED:
+		break;
+	case SEG6_IPTUN_MODE_L2ENCAP_RED:
+		break;
 	default:
 		return -EINVAL;
 	}