[PATCH net v3 0/2] af_unix: Fix scm_stat.nr_fds leak in unix_dgram_disconnected()

[PATCH net v3 0/2] af_unix: Fix scm_stat.nr_fds leak in unix_dgram_disconnected()

[Siddharth Chintamaneni]

This series fixes a u->scm_stat.nr_fds accounting leak in
unix_dgram_disconnected() when purging queued skbs from
sk->sk_receive_queue on disconnect, along with a selftest:

- Patch 1 splices sk->sk_receive_queue onto a local list in
  unix_dgram_disconnected() and calls scm_stat_del() before freeing each
  purged skb.
- Patch 2 adds dgram_disconnect_fdinfo to scm_rights.c verifying scm_fds
  in /proc/self/fdinfo/<fd> across connect(AF_UNSPEC) and dead-peer
  send() disconnects.

Changes in v3:
- Drop unnecessary count_sockets() and sleep(1) calls in
  TEST(dgram_disconnect_fdinfo).
- Link to v2: https://lore.kernel.org/all/20261111225148.88506-1-sidchintamaneni@gmail.com/

Changes in v2:
- Drop MSG_OOB patches and tests per Kuniyuki Iwashima.
- Splice sk->sk_receive_queue onto a local list under the queue lock
  before freeing skbs so concurrent senders cannot extend the drain loop.
- Make dgram_disconnect_fdinfo a standalone TEST() instead of running
  no-op fixture variants, and also test dead-peer send() disconnect.
- Link to v1: https://lore.kernel.org/all/20261106051905.218638-1-sidchintamaneni@gmail.com/

Siddharth Chintamaneni (2):
  af_unix: Fix scm_stat.nr_fds leak in unix_dgram_disconnected().
  selftest: af_unix: Test scm_fds on dgram disconnect.

 net/unix/af_unix.c                            | 59 +++++++++------
 .../selftests/net/af_unix/scm_rights.c        | 75 +++++++++++++++++++
 2 files changed, 110 insertions(+), 24 deletions(-)

-- 
2.43.0

[PATCH net v3 1/2] af_unix: Fix scm_stat.nr_fds leak in unix_dgram_disconnected().

[Siddharth Chintamaneni]

When a connected SOCK_DGRAM socket disconnects (e.g. via connect() with
AF_UNSPEC), reconnects to a different peer, or observes a dead peer in
unix_dgram_sendmsg(), unix_dgram_disconnected() purges all unread
datagrams from sk->sk_receive_queue via skb_queue_purge_reason().

Because skb_queue_purge_reason() frees each skb without calling
scm_stat_del(sk, skb), any SCM_RIGHTS file descriptors attached to the
purged skbs remain permanently accounted in u->scm_stat.nr_fds (reported
as scm_fds in /proc/<pid>/fdinfo/<fd>) for the remaining lifetime of the
socket.

Move scm_stat_add() and scm_stat_del() above unix_dgram_disconnected(),
splice sk->sk_receive_queue onto a local list under the queue lock, and
call scm_stat_del() before kfree_skb_reason() on each purged skb.

Fixes: 3c32da19a858 ("unix: Show number of pending scm files of receive queue in fdinfo")
Signed-off-by: Siddharth Chintamaneni <sidchintamaneni@gmail.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
---
 net/unix/af_unix.c | 59 +++++++++++++++++++++++++++-------------------
 1 file changed, 35 insertions(+), 24 deletions(-)

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 98fcfd7301f6..990a98330d9c 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -609,6 +609,28 @@ static void unix_write_space(struct sock *sk)
 	rcu_read_unlock();
 }
 
+static void scm_stat_add(struct sock *sk, struct sk_buff *skb)
+{
+	struct scm_fp_list *fp = UNIXCB(skb).fp;
+	struct unix_sock *u = unix_sk(sk);
+
+	if (unlikely(fp && fp->count)) {
+		atomic_add(fp->count, &u->scm_stat.nr_fds);
+		unix_add_edges(fp, u);
+	}
+}
+
+static void scm_stat_del(struct sock *sk, struct sk_buff *skb)
+{
+	struct scm_fp_list *fp = UNIXCB(skb).fp;
+	struct unix_sock *u = unix_sk(sk);
+
+	if (unlikely(fp && fp->count)) {
+		atomic_sub(fp->count, &u->scm_stat.nr_fds);
+		unix_del_edges(fp);
+	}
+}
+
 /* When dgram socket disconnects (or changes its peer), we clear its receive
  * queue of packets arrived from previous peer. First, it allows to do
  * flow control based only on wmem_alloc; second, sk connected to peer
@@ -616,8 +638,19 @@ static void unix_write_space(struct sock *sk)
 static void unix_dgram_disconnected(struct sock *sk, struct sock *other)
 {
 	if (!skb_queue_empty(&sk->sk_receive_queue)) {
-		skb_queue_purge_reason(&sk->sk_receive_queue,
-				       SKB_DROP_REASON_UNIX_DISCONNECT);
+		struct sk_buff_head tmp;
+		struct sk_buff *skb;
+
+		__skb_queue_head_init(&tmp);
+
+		spin_lock(&sk->sk_receive_queue.lock);
+		skb_queue_splice_init(&sk->sk_receive_queue, &tmp);
+		spin_unlock(&sk->sk_receive_queue.lock);
+
+		while ((skb = __skb_dequeue(&tmp)) != NULL) {
+			scm_stat_del(sk, skb);
+			kfree_skb_reason(skb, SKB_DROP_REASON_UNIX_DISCONNECT);
+		}
 
 		wake_up_interruptible_all(&unix_sk(sk)->peer_wait);
 
@@ -2058,28 +2091,6 @@ static bool unix_skb_scm_eq(const struct sock *sk, struct sk_buff *skb,
 	       unix_secdata_eq(scm, skb);
 }
 
-static void scm_stat_add(struct sock *sk, struct sk_buff *skb)
-{
-	struct scm_fp_list *fp = UNIXCB(skb).fp;
-	struct unix_sock *u = unix_sk(sk);
-
-	if (unlikely(fp && fp->count)) {
-		atomic_add(fp->count, &u->scm_stat.nr_fds);
-		unix_add_edges(fp, u);
-	}
-}
-
-static void scm_stat_del(struct sock *sk, struct sk_buff *skb)
-{
-	struct scm_fp_list *fp = UNIXCB(skb).fp;
-	struct unix_sock *u = unix_sk(sk);
-
-	if (unlikely(fp && fp->count)) {
-		atomic_sub(fp->count, &u->scm_stat.nr_fds);
-		unix_del_edges(fp);
-	}
-}
-
 static void unix_orphan_scm(struct sock *sk, struct sk_buff *skb)
 {
 	scm_stat_del(sk, skb);
-- 
2.43.0

[PATCH net v3 2/2] selftest: af_unix: Test scm_fds on dgram disconnect.

[Siddharth Chintamaneni]

Add dgram_disconnect_fdinfo in scm_rights.c verifying that scm_fds in
/proc/self/fdinfo/<fd> returns to 0 when a connected SOCK_DGRAM socket
with queued SCM_RIGHTS datagrams disconnects via connect(AF_UNSPEC) or
observes a dead peer in send().

Signed-off-by: Siddharth Chintamaneni <sidchintamaneni@gmail.com>
---
 .../selftests/net/af_unix/scm_rights.c        | 75 +++++++++++++++++++
 1 file changed, 75 insertions(+)

diff --git a/tools/testing/selftests/net/af_unix/scm_rights.c b/tools/testing/selftests/net/af_unix/scm_rights.c
index c165f250220a..19aa26c60f3b 100644
--- a/tools/testing/selftests/net/af_unix/scm_rights.c
+++ b/tools/testing/selftests/net/af_unix/scm_rights.c
@@ -395,4 +395,79 @@ TEST_F(scm_rights, mixed_lowpoint)
 	close_sockets(6);
 }
 
+static int count_scm_fds(struct __test_metadata *_metadata, int fd)
+{
+	char path[64], *line = NULL;
+	int scm_fds = -1, ret;
+	size_t unused = 0;
+	FILE *f;
+
+	snprintf(path, sizeof(path), "/proc/self/fdinfo/%d", fd);
+
+	f = fopen(path, "r");
+	ASSERT_NE(NULL, f);
+
+	while (getline(&line, &unused, f) != -1) {
+		if (strncmp(line, "scm_fds:", 8))
+			continue;
+
+		ret = sscanf(line + 8, "%d", &scm_fds);
+		ASSERT_EQ(1, ret);
+
+		break;
+	}
+
+	free(line);
+
+	ret = fclose(f);
+	ASSERT_EQ(0, ret);
+
+	return scm_fds;
+}
+
+TEST(dgram_disconnect_fdinfo)
+{
+	const FIXTURE_VARIANT(scm_rights) variant = {
+		.name = "UNIX ",
+		.type = SOCK_DGRAM,
+	};
+	FIXTURE_DATA(scm_rights) self = {};
+	struct sockaddr addr = {
+		.sa_family = AF_UNSPEC,
+	};
+	int ret;
+
+	ret = unshare(CLONE_NEWNET);
+	ASSERT_EQ(0, ret);
+
+	create_socketpairs(_metadata, &self, &variant, 2);
+
+	/* 1. Disconnect via connect(AF_UNSPEC) */
+	__send_fd(_metadata, &self, &variant, 0, 0);
+	ASSERT_EQ(2, count_scm_fds(_metadata, self.fd[0]));
+
+	ret = connect(self.fd[0], &addr, sizeof(addr.sa_family));
+	ASSERT_EQ(0, ret);
+	ASSERT_EQ(0, count_scm_fds(_metadata, self.fd[0]));
+
+	/* 2. Dead-peer disconnect via send() */
+	__send_fd(_metadata, &self, &variant, 1, 1);
+	ASSERT_EQ(2, count_scm_fds(_metadata, self.fd[2]));
+
+	ret = close(self.fd[3]);
+	ASSERT_EQ(0, ret);
+
+	ret = send(self.fd[2], "x", 1, 0);
+	ASSERT_EQ(-1, ret);
+	ASSERT_EQ(ECONNREFUSED, errno);
+	ASSERT_EQ(0, count_scm_fds(_metadata, self.fd[2]));
+
+	ret = close(self.fd[0]);
+	ASSERT_EQ(0, ret);
+	ret = close(self.fd[1]);
+	ASSERT_EQ(0, ret);
+	ret = close(self.fd[2]);
+	ASSERT_EQ(0, ret);
+}
+
 TEST_HARNESS_MAIN
-- 
2.43.0