xfrm: Preserve vlan tags for transport mode software GRO

The software GRO path for esp transport mode uses skb_mac_header_rebuild
prior to re-injecting the packet via the xfrm_napi_dev.  This only
copies skb->mac_len bytes of header which may not be sufficient if the
packet contains 802.1Q tags or other VLAN tags.  Worse copying only the
initial header will leave a packet marked as being VLAN tagged but
without the corresponding tag leading to mangling when it is later
untagged.

The VLAN tags are important when receiving the decrypted esp transport
mode packet after GRO processing to ensure it is received on the correct
interface.

Therefore record the full mac header length in xfrm*_transport_input for
later use in corresponding xfrm*_transport_finish to copy the entire mac
header when rebuilding the mac header for GRO.  The skb->data pointer is
left pointing skb->mac_header bytes after the start of the mac header as
is expected by the network stack and network and transport header
offsets reset to this location.

Fixes: 7785bba299a8 ("esp: Add a software GRO codepath")
Signed-off-by: Paul Davey <paul.davey@alliedtelesis.co.nz>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>

5 files changed, 36 insertions, 2 deletions

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 9d24aec064e888..4ff48eda3f642d 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -3031,6 +3031,21 @@ static inline void skb_mac_header_rebuild(struct sk_buff *skb)
 	}
 }
 
+/* Move the full mac header up to current network_header.
+ * Leaves skb->data pointing at offset skb->mac_len into the mac_header.
+ * Must be provided the complete mac header length.
+ */
+static inline void skb_mac_header_rebuild_full(struct sk_buff *skb, u32 full_mac_len)
+{
+	if (skb_mac_header_was_set(skb)) {
+		const unsigned char *old_mac = skb_mac_header(skb);
+
+		skb_set_mac_header(skb, -full_mac_len);
+		memmove(skb_mac_header(skb), old_mac, full_mac_len);
+		__skb_push(skb, full_mac_len - skb->mac_len);
+	}
+}
+
 static inline int skb_checksum_start_offset(const struct sk_buff *skb)
 {
 	return skb->csum_start - skb_headroom(skb);
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 57c743b7e4fef1..cb4841a9fffd27 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -1049,6 +1049,9 @@ struct xfrm_offload {
 #define CRYPTO_INVALID_PACKET_SYNTAX		64
 #define CRYPTO_INVALID_PROTOCOL			128
 
+	/* Used to keep whole l2 header for transport mode GRO */
+	__u32			orig_mac_len;
+
 	__u8			proto;
 	__u8			inner_ipproto;
 };
diff --git a/net/ipv4/xfrm4_input.c b/net/ipv4/xfrm4_input.c
index dae35101d18959..86382e08140ee0 100644
--- a/net/ipv4/xfrm4_input.c
+++ b/net/ipv4/xfrm4_input.c
@@ -63,7 +63,11 @@ int xfrm4_transport_finish(struct sk_buff *skb, int async)
 	ip_send_check(iph);
 
 	if (xo && (xo->flags & XFRM_GRO)) {
-		skb_mac_header_rebuild(skb);
+		/* The full l2 header needs to be preserved so that re-injecting the packet at l2
+		 * works correctly in the presence of vlan tags.
+		 */
+		skb_mac_header_rebuild_full(skb, xo->orig_mac_len);
+		skb_reset_network_header(skb);
 		skb_reset_transport_header(skb);
 		return 0;
 	}
diff --git a/net/ipv6/xfrm6_input.c b/net/ipv6/xfrm6_input.c
index a17d783dc7c0d7..c6b8e132e10a32 100644
--- a/net/ipv6/xfrm6_input.c
+++ b/net/ipv6/xfrm6_input.c
@@ -58,7 +58,11 @@ int xfrm6_transport_finish(struct sk_buff *skb, int async)
 	skb_postpush_rcsum(skb, skb_network_header(skb), nhlen);
 
 	if (xo && (xo->flags & XFRM_GRO)) {
-		skb_mac_header_rebuild(skb);
+		/* The full l2 header needs to be preserved so that re-injecting the packet at l2
+		 * works correctly in the presence of vlan tags.
+		 */
+		skb_mac_header_rebuild_full(skb, xo->orig_mac_len);
+		skb_reset_network_header(skb);
 		skb_reset_transport_header(skb);
 		return 0;
 	}
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index 161f535c8b9495..3a2982a72a6b8c 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -389,11 +389,15 @@ static int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
  */
 static int xfrm4_transport_input(struct xfrm_state *x, struct sk_buff *skb)
 {
+	struct xfrm_offload *xo = xfrm_offload(skb);
 	int ihl = skb->data - skb_transport_header(skb);
 
 	if (skb->transport_header != skb->network_header) {
 		memmove(skb_transport_header(skb),
 			skb_network_header(skb), ihl);
+		if (xo)
+			xo->orig_mac_len =
+				skb_mac_header_was_set(skb) ? skb_mac_header_len(skb) : 0;
 		skb->network_header = skb->transport_header;
 	}
 	ip_hdr(skb)->tot_len = htons(skb->len + ihl);
@@ -404,11 +408,15 @@ static int xfrm4_transport_input(struct xfrm_state *x, struct sk_buff *skb)
 static int xfrm6_transport_input(struct xfrm_state *x, struct sk_buff *skb)
 {
 #if IS_ENABLED(CONFIG_IPV6)
+	struct xfrm_offload *xo = xfrm_offload(skb);
 	int ihl = skb->data - skb_transport_header(skb);
 
 	if (skb->transport_header != skb->network_header) {
 		memmove(skb_transport_header(skb),
 			skb_network_header(skb), ihl);
+		if (xo)
+			xo->orig_mac_len =
+				skb_mac_header_was_set(skb) ? skb_mac_header_len(skb) : 0;
 		skb->network_header = skb->transport_header;
 	}
 	ipv6_hdr(skb)->payload_len = htons(skb->len + ihl -