af_unix: dont send SCM_CREDENTIALS by default

Since commit 7361c36c5224 (af_unix: Allow credentials to work across
user and pid namespaces) af_unix performance dropped a lot.

This is because we now take a reference on pid and cred in each write(),
and release them in read(), usually done from another process,
eventually from another cpu. This triggers false sharing.

# Events: 154K cycles
#
# Overhead  Command       Shared Object        Symbol
# ........  .......  ..................  .........................
#
    10.40%  hackbench  [kernel.kallsyms]   [k] put_pid
     8.60%  hackbench  [kernel.kallsyms]   [k] unix_stream_recvmsg
     7.87%  hackbench  [kernel.kallsyms]   [k] unix_stream_sendmsg
     6.11%  hackbench  [kernel.kallsyms]   [k] do_raw_spin_lock
     4.95%  hackbench  [kernel.kallsyms]   [k] unix_scm_to_skb
     4.87%  hackbench  [kernel.kallsyms]   [k] pid_nr_ns
     4.34%  hackbench  [kernel.kallsyms]   [k] cred_to_ucred
     2.39%  hackbench  [kernel.kallsyms]   [k] unix_destruct_scm
     2.24%  hackbench  [kernel.kallsyms]   [k] sub_preempt_count
     1.75%  hackbench  [kernel.kallsyms]   [k] fget_light
     1.51%  hackbench  [kernel.kallsyms]   [k]
__mutex_lock_interruptible_slowpath
     1.42%  hackbench  [kernel.kallsyms]   [k] sock_alloc_send_pskb

This patch includes SCM_CREDENTIALS information in a af_unix message/skb
only if requested by the sender, [man 7 unix for details how to include
ancillary data using sendmsg() system call]

Note: This might break buggy applications that expected SCM_CREDENTIAL
from an unaware write() system call, and receiver not using SO_PASSCRED
socket option.

If SOCK_PASSCRED is set on source or destination socket, we still
include credentials for mere write() syscalls.

Performance boost in hackbench : more than 50% gain on a 16 thread
machine (2 quad-core cpus, 2 threads per core)

hackbench 20 thread 2000

4.228 sec instead of 9.102 sec

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 23 insertions, 1 deletions

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index ec68e1c05b8..466fbcc5cf7 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1381,8 +1381,10 @@ static int unix_attach_fds(struct scm_cookie *scm, struct sk_buff *skb)
 static int unix_scm_to_skb(struct scm_cookie *scm, struct sk_buff *skb, bool send_fds)
 {
 	int err = 0;
+
 	UNIXCB(skb).pid  = get_pid(scm->pid);
-	UNIXCB(skb).cred = get_cred(scm->cred);
+	if (scm->cred)
+		UNIXCB(skb).cred = get_cred(scm->cred);
 	UNIXCB(skb).fp = NULL;
 	if (scm->fp && send_fds)
 		err = unix_attach_fds(scm, skb);
@@ -1392,6 +1394,24 @@ static int unix_scm_to_skb(struct scm_cookie *scm, struct sk_buff *skb, bool sen
 }
 
 /*
+ * Some apps rely on write() giving SCM_CREDENTIALS
+ * We include credentials if source or destination socket
+ * asserted SOCK_PASSCRED.
+ */
+static void maybe_add_creds(struct sk_buff *skb, const struct socket *sock,
+			    const struct sock *other)
+{
+	if (UNIXCB(skb).cred)
+		return;
+	if (test_bit(SOCK_PASSCRED, &sock->flags) ||
+	    !other->sk_socket ||
+	    test_bit(SOCK_PASSCRED, &other->sk_socket->flags)) {
+		UNIXCB(skb).pid  = get_pid(task_tgid(current));
+		UNIXCB(skb).cred = get_current_cred();
+	}
+}
+
+/*
  *	Send AF_UNIX data.
  */
 
@@ -1538,6 +1558,7 @@ restart:
 
 	if (sock_flag(other, SOCK_RCVTSTAMP))
 		__net_timestamp(skb);
+	maybe_add_creds(skb, sock, other);
 	skb_queue_tail(&other->sk_receive_queue, skb);
 	if (max_level > unix_sk(other)->recursion_level)
 		unix_sk(other)->recursion_level = max_level;
@@ -1652,6 +1673,7 @@ static int unix_stream_sendmsg(struct kiocb *kiocb, struct socket *sock,
 		    (other->sk_shutdown & RCV_SHUTDOWN))
 			goto pipe_err_free;
 
+		maybe_add_creds(skb, sock, other);
 		skb_queue_tail(&other->sk_receive_queue, skb);
 		if (max_level > unix_sk(other)->recursion_level)
 			unix_sk(other)->recursion_level = max_level;