bpf: add bpf_skb_adjust_room mode BPF_ADJ_ROOM_MAC

bpf_skb_adjust_room net allows inserting room in an skb.

Existing mode BPF_ADJ_ROOM_NET inserts room after the network header
by pulling the skb, moving the network header forward and zeroing the
new space.

Add new mode BPF_ADJUST_ROOM_MAC that inserts room after the mac
header. This allows inserting tunnel headers in front of the network
header without having to recreate the network header in the original
space, avoiding two copies.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

include/uapi/linux/bpf.h

@@ -1478,7 +1478,10 @@ union bpf_attr {
  * 		Grow or shrink the room for data in the packet associated to
  * 		*skb* by *len_diff*, and according to the selected *mode*.
  *
- * 		There is a single supported mode at this time:
+ *		There are two supported modes at this time:
+ *
+ *		* **BPF_ADJ_ROOM_MAC**: Adjust room at the mac layer
+ *		  (room space is added or removed below the layer 2 header).
  *
  * 		* **BPF_ADJ_ROOM_NET**: Adjust room at the network layer
  * 		  (room space is added or removed below the layer 3 header).
@@ -2627,6 +2630,7 @@ enum bpf_func_id {
 /* Mode for BPF_FUNC_skb_adjust_room helper. */
 enum bpf_adj_room_mode {
 	BPF_ADJ_ROOM_NET,
+	BPF_ADJ_ROOM_MAC,
 };
 
 /* Mode for BPF_FUNC_skb_load_bytes_relative helper. */

net/core/filter.c

@@ -2963,9 +2963,8 @@ static u32 bpf_skb_net_base_len(const struct sk_buff *skb)
 	}
 }
 
-static int bpf_skb_net_grow(struct sk_buff *skb, u32 len_diff)
+static int bpf_skb_net_grow(struct sk_buff *skb, u32 off, u32 len_diff)
 {
-	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
 	int ret;
 
 	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
@@ -2992,9 +2991,8 @@ static int bpf_skb_net_grow(struct sk_buff *skb, u32 len_diff)
 	return 0;
 }
 
-static int bpf_skb_net_shrink(struct sk_buff *skb, u32 len_diff)
+static int bpf_skb_net_shrink(struct sk_buff *skb, u32 off, u32 len_diff)
 {
-	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
 	int ret;
 
 	if (skb_is_gso(skb) && !skb_is_gso_tcp(skb))
@@ -3027,7 +3025,8 @@ static u32 __bpf_skb_max_len(const struct sk_buff *skb)
 			  SKB_MAX_ALLOC;
 }
 
-static int bpf_skb_adjust_net(struct sk_buff *skb, s32 len_diff)
+BPF_CALL_4(bpf_skb_adjust_room, struct sk_buff *, skb, s32, len_diff,
+	   u32, mode, u64, flags)
 {
 	bool trans_same = skb->transport_header == skb->network_header;
 	u32 len_cur, len_diff_abs = abs(len_diff);
@@ -3035,14 +3034,28 @@ static int bpf_skb_adjust_net(struct sk_buff *skb, s32 len_diff)
 	u32 len_max = __bpf_skb_max_len(skb);
 	__be16 proto = skb->protocol;
 	bool shrink = len_diff < 0;
+	u32 off;
 	int ret;
 
+	if (unlikely(flags))
+		return -EINVAL;
 	if (unlikely(len_diff_abs > 0xfffU))
 		return -EFAULT;
 	if (unlikely(proto != htons(ETH_P_IP) &&
 		     proto != htons(ETH_P_IPV6)))
 		return -ENOTSUPP;
 
+	off = skb_mac_header_len(skb);
+	switch (mode) {
+	case BPF_ADJ_ROOM_NET:
+		off += bpf_skb_net_base_len(skb);
+		break;
+	case BPF_ADJ_ROOM_MAC:
+		break;
+	default:
+		return -ENOTSUPP;
+	}
+
 	len_cur = skb->len - skb_network_offset(skb);
 	if (skb_transport_header_was_set(skb) && !trans_same)
 		len_cur = skb_network_header_len(skb);
@@ -3052,24 +3065,13 @@ static int bpf_skb_adjust_net(struct sk_buff *skb, s32 len_diff)
 			 !skb_is_gso(skb))))
 		return -ENOTSUPP;
 
-	ret = shrink ? bpf_skb_net_shrink(skb, len_diff_abs) :
-		       bpf_skb_net_grow(skb, len_diff_abs);
+	ret = shrink ? bpf_skb_net_shrink(skb, off, len_diff_abs) :
+		       bpf_skb_net_grow(skb, off, len_diff_abs);
 
 	bpf_compute_data_pointers(skb);
 	return ret;
 }
 
-BPF_CALL_4(bpf_skb_adjust_room, struct sk_buff *, skb, s32, len_diff,
-	   u32, mode, u64, flags)
-{
-	if (unlikely(flags))
-		return -EINVAL;
-	if (likely(mode == BPF_ADJ_ROOM_NET))
-		return bpf_skb_adjust_net(skb, len_diff);
-
-	return -ENOTSUPP;
-}
-
 static const struct bpf_func_proto bpf_skb_adjust_room_proto = {
 	.func		= bpf_skb_adjust_room,
 	.gpl_only	= false,