In the error path of netlink_proto_init(), frees the already allocated
bucket table for new hash tables in a loop, but it is going to panic,
so it is not necessary to clean up the resources, just remove the
dead code.

Suggested-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
v2:
- Add Suggested-by.
- Remove the fix tag.
- Remove instead of check >=0.
- Update the commit message.
---
 net/netlink/af_netlink.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 0a9287fadb47..52a7c7233cab 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -2935,12 +2935,8 @@ static int __init netlink_proto_init(void)
 
 	for (i = 0; i < MAX_LINKS; i++) {
 		if (rhashtable_init(&nl_table[i].hash,
-				    &netlink_rhashtable_params) < 0) {
-			while (--i > 0)
-				rhashtable_destroy(&nl_table[i].hash);
-			kfree(nl_table);
+				    &netlink_rhashtable_params) < 0)
 			goto panic;
-		}
 	}
 
 	netlink_add_usersock_entry();
-- 
2.34.1