bpf: add bpf_skb_adjust_room helper

This work adds a helper that can be used to adjust net room of an
skb. The helper is generic and can be further extended in future.
Main use case is for having a programmatic way to add/remove room to
v4/v6 header options along with cls_bpf on egress and ingress hook
of the data path. It reuses most of the infrastructure that we added
for the bpf_skb_change_type() helper which can be used in nat64
translations. Similarly, the helper only takes care of adjusting the
room so that related data is populated and csum adapted out of the
BPF program using it.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/uapi/linux/bpf.h

@@ -531,6 +531,14 @@ union bpf_attr {
  *     @optval: pointer to option value
  *     @optlen: length of optval in byes
  *     Return: 0 or negative error
+ *
+ * int bpf_skb_adjust_room(skb, len_diff, mode, flags)
+ *     Grow or shrink room in sk_buff.
+ *     @skb: pointer to skb
+ *     @len_diff: (signed) amount of room to grow/shrink
+ *     @mode: operation mode (enum bpf_adj_room_mode)
+ *     @flags: reserved for future use
+ *     Return: 0 on success or negative error code
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -582,7 +590,8 @@ union bpf_attr {
 	FN(get_socket_cookie),		\
 	FN(get_socket_uid),		\
 	FN(set_hash),			\
-	FN(setsockopt),
+	FN(setsockopt),			\
+	FN(skb_adjust_room),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
@@ -632,6 +641,11 @@ enum bpf_func_id {
 /* BPF_FUNC_perf_event_output for sk_buff input context. */
 #define BPF_F_CTXLEN_MASK		(0xfffffULL << 32)
 
+/* Mode for BPF_FUNC_skb_adjust_room helper. */
+enum bpf_adj_room_mode {
+	BPF_ADJ_ROOM_NET,
+};
+
 /* user accessible mirror of in-kernel sk_buff.
  * new fields can only be added to the end of this structure
  */

net/core/filter.c

@@ -2154,6 +2154,124 @@ static const struct bpf_func_proto bpf_skb_change_type_proto = {
 	.arg2_type	= ARG_ANYTHING,
 };
 
+static u32 bpf_skb_net_base_len(const struct sk_buff *skb)
+{
+	switch (skb->protocol) {
+	case htons(ETH_P_IP):
+		return sizeof(struct iphdr);
+	case htons(ETH_P_IPV6):
+		return sizeof(struct ipv6hdr);
+	default:
+		return ~0U;
+	}
+}
+
+static int bpf_skb_net_grow(struct sk_buff *skb, u32 len_diff)
+{
+	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
+	int ret;
+
+	ret = skb_cow(skb, len_diff);
+	if (unlikely(ret < 0))
+		return ret;
+
+	ret = bpf_skb_net_hdr_push(skb, off, len_diff);
+	if (unlikely(ret < 0))
+		return ret;
+
+	if (skb_is_gso(skb)) {
+		/* Due to header grow, MSS needs to be downgraded. */
+		skb_shinfo(skb)->gso_size -= len_diff;
+		/* Header must be checked, and gso_segs recomputed. */
+		skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
+		skb_shinfo(skb)->gso_segs = 0;
+	}
+
+	return 0;
+}
+
+static int bpf_skb_net_shrink(struct sk_buff *skb, u32 len_diff)
+{
+	u32 off = skb_mac_header_len(skb) + bpf_skb_net_base_len(skb);
+	int ret;
+
+	ret = skb_unclone(skb, GFP_ATOMIC);
+	if (unlikely(ret < 0))
+		return ret;
+
+	ret = bpf_skb_net_hdr_pop(skb, off, len_diff);
+	if (unlikely(ret < 0))
+		return ret;
+
+	if (skb_is_gso(skb)) {
+		/* Due to header shrink, MSS can be upgraded. */
+		skb_shinfo(skb)->gso_size += len_diff;
+		/* Header must be checked, and gso_segs recomputed. */
+		skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
+		skb_shinfo(skb)->gso_segs = 0;
+	}
+
+	return 0;
+}
+
+static u32 __bpf_skb_max_len(const struct sk_buff *skb)
+{
+	return skb->dev->mtu + skb->dev->hard_header_len;
+}
+
+static int bpf_skb_adjust_net(struct sk_buff *skb, s32 len_diff)
+{
+	bool trans_same = skb->transport_header == skb->network_header;
+	u32 len_cur, len_diff_abs = abs(len_diff);
+	u32 len_min = bpf_skb_net_base_len(skb);
+	u32 len_max = __bpf_skb_max_len(skb);
+	__be16 proto = skb->protocol;
+	bool shrink = len_diff < 0;
+	int ret;
+
+	if (unlikely(len_diff_abs > 0xfffU))
+		return -EFAULT;
+	if (unlikely(proto != htons(ETH_P_IP) &&
+		     proto != htons(ETH_P_IPV6)))
+		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);
+	if ((shrink && (len_diff_abs >= len_cur ||
+			len_cur - len_diff_abs < len_min)) ||
+	    (!shrink && (skb->len + len_diff_abs > len_max &&
+			 !skb_is_gso(skb))))
+		return -ENOTSUPP;
+
+	ret = shrink ? bpf_skb_net_shrink(skb, len_diff_abs) :
+		       bpf_skb_net_grow(skb, len_diff_abs);
+
+	bpf_compute_data_end(skb);
+	return 0;
+}
+
+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,
+	.ret_type	= RET_INTEGER,
+	.arg1_type	= ARG_PTR_TO_CTX,
+	.arg2_type	= ARG_ANYTHING,
+	.arg3_type	= ARG_ANYTHING,
+	.arg4_type	= ARG_ANYTHING,
+};
+
 static u32 __bpf_skb_min_len(const struct sk_buff *skb)
 {
 	u32 min_len = skb_network_offset(skb);
@@ -2166,11 +2284,6 @@ static u32 __bpf_skb_min_len(const struct sk_buff *skb)
 	return min_len;
 }
 
-static u32 __bpf_skb_max_len(const struct sk_buff *skb)
-{
-	return skb->dev->mtu + skb->dev->hard_header_len;
-}
-
 static int bpf_skb_grow_rcsum(struct sk_buff *skb, unsigned int new_len)
 {
 	unsigned int old_len = skb->len;
@@ -2307,6 +2420,7 @@ bool bpf_helper_changes_pkt_data(void *func)
 	    func == bpf_skb_change_proto ||
 	    func == bpf_skb_change_head ||
 	    func == bpf_skb_change_tail ||
+	    func == bpf_skb_adjust_room ||
 	    func == bpf_skb_pull_data ||
 	    func == bpf_clone_redirect ||
 	    func == bpf_l3_csum_replace ||
@@ -2849,6 +2963,8 @@ tc_cls_act_func_proto(enum bpf_func_id func_id)
 		return &bpf_skb_change_proto_proto;
 	case BPF_FUNC_skb_change_type:
 		return &bpf_skb_change_type_proto;
+	case BPF_FUNC_skb_adjust_room:
+		return &bpf_skb_adjust_room_proto;
 	case BPF_FUNC_skb_change_tail:
 		return &bpf_skb_change_tail_proto;
 	case BPF_FUNC_skb_get_tunnel_key:

tools/include/uapi/linux/bpf.h

@@ -531,6 +531,14 @@ union bpf_attr {
  *     @optval: pointer to option value
  *     @optlen: length of optval in byes
  *     Return: 0 or negative error
+ *
+ * int bpf_skb_adjust_room(skb, len_diff, mode, flags)
+ *     Grow or shrink room in sk_buff.
+ *     @skb: pointer to skb
+ *     @len_diff: (signed) amount of room to grow/shrink
+ *     @mode: operation mode (enum bpf_adj_room_mode)
+ *     @flags: reserved for future use
+ *     Return: 0 on success or negative error code
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -582,7 +590,8 @@ union bpf_attr {
 	FN(get_socket_cookie),		\
 	FN(get_socket_uid),		\
 	FN(set_hash),			\
-	FN(setsockopt),
+	FN(setsockopt),			\
+	FN(skb_adjust_room),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
@@ -632,6 +641,11 @@ enum bpf_func_id {
 /* BPF_FUNC_perf_event_output for sk_buff input context. */
 #define BPF_F_CTXLEN_MASK		(0xfffffULL << 32)
 
+/* Mode for BPF_FUNC_skb_adjust_room helper. */
+enum bpf_adj_room_mode {
+	BPF_ADJ_ROOM_NET_OPTS,
+};
+
 /* user accessible mirror of in-kernel sk_buff.
  * new fields can only be added to the end of this structure
  */