[4/5] sock_poll_wait: kill the no longer necessary barrier after poll_wait()

Commit Message

Oleg Nesterov Jan. 7, 2025, 4:27 p.m. UTC

Now that poll_wait() provides a full barrier we can remove smp_mb() from
sock_poll_wait().

Also, the poll_does_not_wait() check before poll_wait() just adds the
unnecessary confusion, kill it. poll_wait() does the same "p && p->_qproc"
check.

Signed-off-by: Oleg Nesterov <[e-mail]>
---
 include/net/sock.h | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

Patch

diff --git a/include/net/sock.h b/include/net/sock.h
index 7464e9f9f47c..305f3ae5edc2 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -2291,7 +2291,7 @@  static inline bool skwq_has_sleeper(struct socket_wq *wq)
 }
 
 /**
- * sock_poll_wait - place memory barrier behind the poll_wait call.
+ * sock_poll_wait - wrapper for the poll_wait call.
  * @filp:           file
  * @sock:           socket to wait on
  * @p:              poll_table
@@ -2301,15 +2301,12 @@  static inline bool skwq_has_sleeper(struct socket_wq *wq)
 static inline void sock_poll_wait(struct file *filp, struct socket *sock,
 				  poll_table *p)
 {
-	if (!poll_does_not_wait(p)) {
-		poll_wait(filp, &sock->wq.wait, p);
-		/* We need to be sure we are in sync with the
-		 * socket flags modification.
-		 *
-		 * This memory barrier is paired in the wq_has_sleeper.
-		 */
-		smp_mb();
-	}
+	/* Provides a barrier we need to be sure we are in sync
+	 * with the socket flags modification.
+	 *
+	 * This memory barrier is paired in the wq_has_sleeper.
+	 */
+	poll_wait(filp, &sock->wq.wait, p);
 }
 
 static inline void skb_set_hash_from_sk(struct sk_buff *skb, struct sock *sk)