[Devel] [PATCH RH7] Fix a leak in socket(2) when we fail to allocate a file descriptor.

[Pavel Tikhomirov]

From: Al Viro <viro at zeniv.linux.org.uk>

Got broken by "make sock_alloc_file() do sock_release() on failures" -
cleanup after sock_map_fd() failure got pulled all the way into
sock_alloc_file(), but it used to serve the case when sock_map_fd()
failed *before* getting to sock_alloc_file() as well, and that got
lost.  Trivial to fix, fortunately.

Fixes: 8e1611e23579 (make sock_alloc_file() do sock_release() on failures)
Reported-by: Dmitry Vyukov <dvyukov at google.com>
Signed-off-by: Al Viro <viro at zeniv.linux.org.uk>

We see that lldpad eats all kernel memory on the node with those leaked
sockets constantly looping on SyS_socket with EMFILE error.

https://jira.sw.ru/browse/PSBM-133610
Signed-off-by: Pavel Tikhomirov <ptikhomirov at virtuozzo.com>
---
Note: already have it in VZ8.
---
 net/socket.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/socket.c b/net/socket.c
index e720aa814acc..bcf98c0f2fb9 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -403,8 +403,10 @@ static int sock_map_fd(struct socket *sock, int flags)
 {
 	struct file *newfile;
 	int fd = get_unused_fd_flags(flags);
-	if (unlikely(fd < 0))
+	if (unlikely(fd < 0)) {
+		sock_release(sock);
 		return fd;
+	}
 
 	newfile = sock_alloc_file(sock, flags, NULL);
 	if (likely(!IS_ERR(newfile))) {
-- 
2.31.1