netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kuniyuki Iwashima <kuniyu@amazon.com>
To: "David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>
Cc: Kuniyuki Iwashima <kuniyu@amazon.com>,
	Kuniyuki Iwashima <kuni1840@gmail.com>, <netdev@vger.kernel.org>
Subject: [PATCH v1 net-next 14/15] af_unix: Remove sk_locked logic in unix_dgram_sendmsg().
Date: Fri, 6 Dec 2024 14:26:06 +0900	[thread overview]
Message-ID: <20241206052607.1197-15-kuniyu@amazon.com> (raw)
In-Reply-To: <20241206052607.1197-1-kuniyu@amazon.com>

If the receiver socket is not connected to the sender socket and the
receiver's queue is full in unix_dgram_sendmsg(), we lock the both
sockets and retry.

The logic adds one unlikely check to the sane path and complicates the
entire function, but it's not worth that.

Let's remove the sk_locked variable and simplify the logic and the error
path.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
---
 net/unix/af_unix.c | 29 +++++++++--------------------
 1 file changed, 9 insertions(+), 20 deletions(-)

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 629e7a81178e..594c7428f22d 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1967,7 +1967,6 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 	struct scm_cookie scm;
 	struct sk_buff *skb;
 	int data_len = 0;
-	int sk_locked;
 	long timeo;
 	int err;
 
@@ -2059,7 +2058,6 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 	}
 
 restart:
-	sk_locked = 0;
 	unix_state_lock(other);
 restart_locked:
 
@@ -2082,8 +2080,7 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 			goto out_sock_put;
 		}
 
-		if (!sk_locked)
-			unix_state_lock(sk);
+		unix_state_lock(sk);
 
 		if (unix_peer(sk) == other) {
 			unix_peer(sk) = NULL;
@@ -2136,27 +2133,21 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 			goto restart;
 		}
 
-		if (!sk_locked) {
-			unix_state_unlock(other);
-			unix_state_double_lock(sk, other);
-		}
+		unix_state_unlock(other);
+		unix_state_double_lock(sk, other);
 
 		if (unix_peer(sk) != other ||
-		    unix_dgram_peer_wake_me(sk, other)) {
+		    unix_dgram_peer_wake_me(sk, other))
 			err = -EAGAIN;
-			sk_locked = 1;
+
+		unix_state_unlock(sk);
+
+		if (err)
 			goto out_unlock;
-		}
 
-		if (!sk_locked) {
-			sk_locked = 1;
-			goto restart_locked;
-		}
+		goto restart_locked;
 	}
 
-	if (unlikely(sk_locked))
-		unix_state_unlock(sk);
-
 	if (sock_flag(other, SOCK_RCVTSTAMP))
 		__net_timestamp(skb);
 	maybe_add_creds(skb, sock, other);
@@ -2169,8 +2160,6 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 	return len;
 
 out_unlock:
-	if (sk_locked)
-		unix_state_unlock(sk);
 	unix_state_unlock(other);
 out_sock_put:
 	sock_put(other);
-- 
2.39.5 (Apple Git-154)


  parent reply	other threads:[~2024-12-06  5:31 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-06  5:25 [PATCH v1 net-next 00/15] af_unix: Prepare for skb drop reason Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 01/15] af_unix: Set error only when needed in unix_stream_connect() Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 02/15] af_unix: Clean up error paths " Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 03/15] af_unix: Set error only when needed in unix_stream_sendmsg() Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 04/15] af_unix: Remove redundant SEND_SHUTDOWN check " Kuniyuki Iwashima
2024-12-10 11:48   ` Paolo Abeni
2024-12-13  8:18     ` Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 05/15] af_unix: Clean up error paths " Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 06/15] af_unix: Set error only when needed in unix_dgram_sendmsg() Kuniyuki Iwashima
2024-12-06  5:25 ` [PATCH v1 net-next 07/15] af_unix: Call unix_autobind() only when msg_namelen is specified " Kuniyuki Iwashima
2024-12-10 11:11   ` Paolo Abeni
2024-12-10 11:33     ` Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 08/15] af_unix: Move !sunaddr case " Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 09/15] af_unix: Use msg->{msg_name,msg_namelen} " Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 10/15] af_unix: Split restart label " Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 11/15] af_unix: Defer sock_put() to clean up path " Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 12/15] af_unix: Clean up SOCK_DEAD error paths " Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 13/15] af_unix: Clean up error path " Kuniyuki Iwashima
2024-12-06  5:26 ` Kuniyuki Iwashima [this message]
2024-12-10 11:42   ` [PATCH v1 net-next 14/15] af_unix: Remove sk_locked logic " Paolo Abeni
2024-12-13  8:26     ` Kuniyuki Iwashima
2024-12-06  5:26 ` [PATCH v1 net-next 15/15] af_unix: Remove unix_our_peer() Kuniyuki Iwashima

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20241206052607.1197-15-kuniyu@amazon.com \
    --to=kuniyu@amazon.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kuba@kernel.org \
    --cc=kuni1840@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).