From: Eric Dumazet <edumazet@google.com>
Date: Thu, 20 Mar 2025 12:16:04 +0000 (+0000)
Subject: tcp: avoid atomic operations on sk->sk_rmem_alloc
X-Git-Tag: v6.15-rc1~160^2~25
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=0de2a5c4b824da2205658ebebb99a55c43cdf60f;p=thirdparty%2Fkernel%2Flinux.git

tcp: avoid atomic operations on sk->sk_rmem_alloc

TCP uses generic skb_set_owner_r() and sock_rfree()
for received packets, with socket lock being owned.

Switch to private versions, avoiding two atomic operations
per packet.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250320121604.3342831-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 4450c384ef178..df04dc09c519d 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -779,6 +779,7 @@ static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize)
 
 /* tcp.c */
 void tcp_get_info(struct sock *, struct tcp_info *);
+void tcp_sock_rfree(struct sk_buff *skb);
 
 /* Read 'sendfile()'-style from a TCP socket */
 int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
@@ -2898,4 +2899,18 @@ enum skb_drop_reason tcp_inbound_hash(struct sock *sk,
 		const void *saddr, const void *daddr,
 		int family, int dif, int sdif);
 
+/* version of skb_set_owner_r() avoiding one atomic_add() */
+static inline void tcp_skb_set_owner_r(struct sk_buff *skb, struct sock *sk)
+{
+	skb_orphan(skb);
+	skb->sk = sk;
+	skb->destructor = tcp_sock_rfree;
+
+	sock_owned_by_me(sk);
+	atomic_set(&sk->sk_rmem_alloc,
+		   atomic_read(&sk->sk_rmem_alloc) + skb->truesize);
+
+	sk_forward_alloc_add(sk, -skb->truesize);
+}
+
 #endif	/* _TCP_H */
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 6edc441b37023..ea8de00f669d0 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1525,11 +1525,25 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied)
 	__tcp_cleanup_rbuf(sk, copied);
 }
 
+/* private version of sock_rfree() avoiding one atomic_sub() */
+void tcp_sock_rfree(struct sk_buff *skb)
+{
+	struct sock *sk = skb->sk;
+	unsigned int len = skb->truesize;
+
+	sock_owned_by_me(sk);
+	atomic_set(&sk->sk_rmem_alloc,
+		   atomic_read(&sk->sk_rmem_alloc) - len);
+
+	sk_forward_alloc_add(sk, len);
+	sk_mem_reclaim(sk);
+}
+
 static void tcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb)
 {
 	__skb_unlink(skb, &sk->sk_receive_queue);
-	if (likely(skb->destructor == sock_rfree)) {
-		sock_rfree(skb);
+	if (likely(skb->destructor == tcp_sock_rfree)) {
+		tcp_sock_rfree(skb);
 		skb->destructor = NULL;
 		skb->sk = NULL;
 		return skb_attempt_defer_free(skb);
diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
index 1a6b1bc542451..ca40665145c69 100644
--- a/net/ipv4/tcp_fastopen.c
+++ b/net/ipv4/tcp_fastopen.c
@@ -189,7 +189,7 @@ void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb)
 	tcp_segs_in(tp, skb);
 	__skb_pull(skb, tcp_hdrlen(skb));
 	sk_forced_mem_schedule(sk, skb->truesize);
-	skb_set_owner_r(skb, sk);
+	tcp_skb_set_owner_r(skb, sk);
 
 	TCP_SKB_CB(skb)->seq++;
 	TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_SYN;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index a35018e2d0ba2..e1f952fbac48d 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -5171,7 +5171,7 @@ end:
 		if (tcp_is_sack(tp))
 			tcp_grow_window(sk, skb, false);
 		skb_condense(skb);
-		skb_set_owner_r(skb, sk);
+		tcp_skb_set_owner_r(skb, sk);
 	}
 }
 
@@ -5187,7 +5187,7 @@ static int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb,
 	tcp_rcv_nxt_update(tcp_sk(sk), TCP_SKB_CB(skb)->end_seq);
 	if (!eaten) {
 		tcp_add_receive_queue(sk, skb);
-		skb_set_owner_r(skb, sk);
+		tcp_skb_set_owner_r(skb, sk);
 	}
 	return eaten;
 }
@@ -5504,7 +5504,7 @@ skip_this:
 			__skb_queue_before(list, skb, nskb);
 		else
 			__skb_queue_tail(&tmp, nskb); /* defer rbtree insertion */
-		skb_set_owner_r(nskb, sk);
+		tcp_skb_set_owner_r(nskb, sk);
 		mptcp_skb_ext_move(nskb, skb);
 
 		/* Copy data, releasing collapsed skbs. */