bpf: Don't refcount LISTEN sockets in sk_assign()

Avoid taking a reference on listen sockets by checking the socket type in the sk_assign and in the corresponding skb_steal_sock() code in the the transport layer, and by ensuring that the prefetch free (sock_pfree) function uses the same logic to check whether the socket is refcounted. Suggested-by: Martin KaFai Lau <[email protected]> Signed-off-by: Joe Stringer <[email protected]> Signed-off-by: Alexei Starovoitov <[email protected]> Acked-by: Martin KaFai Lau <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
kernel-patches · Mar 30, 2020 · 7ae215d · 7ae215d
1 parent 71489e2
commit 7ae215d
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 12 deletions.
diff --git a/include/net/sock.h b/include/net/sock.h
@@ -2537,6 +2537,21 @@ skb_sk_is_prefetched(struct sk_buff *skb)
 #endif /* CONFIG_INET */
 }
 
+/* This helper checks if a socket is a full socket,
+ * ie _not_ a timewait or request socket.
+ */
+static inline bool sk_fullsock(const struct sock *sk)
+{
+	return (1 << sk->sk_state) & ~(TCPF_TIME_WAIT | TCPF_NEW_SYN_RECV);
+}
+
+static inline bool
+sk_is_refcounted(struct sock *sk)
+{
+	/* Only full sockets have sk->sk_flags. */
+	return !sk_fullsock(sk) || !sock_flag(sk, SOCK_RCU_FREE);
+}
+
 /**
  * skb_steal_sock
  * @skb to steal the socket from
@@ -2549,6 +2564,8 @@ skb_steal_sock(struct sk_buff *skb, bool *refcounted)
 		struct sock *sk = skb->sk;
 
 		*refcounted = true;
+		if (skb_sk_is_prefetched(skb))
+			*refcounted = sk_is_refcounted(sk);
 		skb->destructor = NULL;
 		skb->sk = NULL;
 		return sk;
@@ -2557,14 +2574,6 @@ skb_steal_sock(struct sk_buff *skb, bool *refcounted)
 	return NULL;
 }
 
-/* This helper checks if a socket is a full socket,
- * ie _not_ a timewait or request socket.
- */
-static inline bool sk_fullsock(const struct sock *sk)
-{
-	return (1 << sk->sk_state) & ~(TCPF_TIME_WAIT | TCPF_NEW_SYN_RECV);
-}
-
 /* Checks if this SKB belongs to an HW offloaded socket
  * and whether any SW fallbacks are required based on dev.
  * Check decrypted mark in case skb_orphan() cleared socket.

diff --git a/net/core/filter.c b/net/core/filter.c
@@ -5401,8 +5401,7 @@ static const struct bpf_func_proto bpf_sk_lookup_udp_proto = {
 
 BPF_CALL_1(bpf_sk_release, struct sock *, sk)
 {
-	/* Only full sockets have sk->sk_flags. */
-	if (!sk_fullsock(sk) || !sock_flag(sk, SOCK_RCU_FREE))
+	if (sk_is_refcounted(sk))
 		sock_gen_put(sk);
 	return 0;
 }
@@ -5928,7 +5927,8 @@ BPF_CALL_3(bpf_sk_assign, struct sk_buff *, skb, struct sock *, sk, u64, flags)
 		return -ENETUNREACH;
 	if (unlikely(sk->sk_reuseport))
 		return -ESOCKTNOSUPPORT;
-	if (unlikely(!refcount_inc_not_zero(&sk->sk_refcnt)))
+	if (sk_is_refcounted(sk) &&
+	    unlikely(!refcount_inc_not_zero(&sk->sk_refcnt)))
 		return -ENOENT;
 
 	skb_orphan(skb);

diff --git a/net/core/sock.c b/net/core/sock.c
@@ -2077,7 +2077,8 @@ EXPORT_SYMBOL(sock_efree);
 #ifdef CONFIG_INET
 void sock_pfree(struct sk_buff *skb)
 {
-	sock_gen_put(skb->sk);
+	if (sk_is_refcounted(skb->sk))
+		sock_gen_put(skb->sk);
 }
 EXPORT_SYMBOL(sock_pfree);
 #endif /* CONFIG_INET */