Re: [PATCH net-next v14 4/9] net: devmem: Implement TX path

From
Paolo Abeni <pabeni@redhat.com>
Date
2025-05-13 09:12:12
Also in
io-uring, kvm, linux-doc, linux-kselftest, lkml, virtualization
Reply to
Mina Almasry, 2025-05-08
On 5/8/25 2:48 AM, Mina Almasry wrote:
quoted hunk ↗ jump to hunk
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 86c427f166367..0ae265d39184e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1059,6 +1059,7 @@ int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, int *copied,
 
 int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 {
+	struct net_devmem_dmabuf_binding *binding = NULL;
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct ubuf_info *uarg = NULL;
 	struct sk_buff *skb;
@@ -1066,11 +1067,23 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 	int flags, err, copied = 0;
 	int mss_now = 0, size_goal, copied_syn = 0;
 	int process_backlog = 0;
+	bool sockc_valid = true;
 	int zc = 0;
 	long timeo;
 
 	flags = msg->msg_flags;
 
+	sockc = (struct sockcm_cookie){ .tsflags = READ_ONCE(sk->sk_tsflags) };
+	if (msg->msg_controllen) {
+		err = sock_cmsg_send(sk, msg, &sockc);
+		if (unlikely(err))
+			/* Don't return error until MSG_FASTOPEN has been
+			 * processed; that may succeed even if the cmsg is
+			 * invalid.
+			 */
+			sockc_valid = false;
It occurred to me a bit too late that this chunk of code could be
cleaned-up a bit using a 'sockc_err' variable to store the
sock_cmsg_send() return code instead of the 'sockc_valid' bool. It
should avoid a conditional here and in the later error check.

(just to mention a possible follow-up! no need to repost!)

Thanks,

Paolo
Thread (12 messages) 12 messages, 3 authors, last reply 2025-05-13
Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help