netdev - [net-next 3/3] tipc: step sk->sk_drops when rcv buffer is full

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 21 Mar 2018 14:37:45 +0100
From:   GhantaKrishnamurthy MohanKrishna 
        <mohan.krishna.ghanta.krishnamurthy@...csson.com>
To:     <tipc-discussion@...ts.sourceforge.net>, <jon.maloy@...csson.com>,
        <maloy@...jonn.com>, <ying.xue@...driver.com>,
        <mohan.krishna.ghanta.krishnamurthy@...csson.com>,
        <netdev@...r.kernel.org>, <davem@...emloft.net>
CC:     Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan@...il.com>
Subject: [net-next  3/3] tipc: step sk->sk_drops when rcv buffer is full

Currently when tipc is unable to queue a received message on a
socket, the message is rejected back to the sender with error
TIPC_ERR_OVERLOAD. However, the application on this socket
has no knowledge about these discards.

In this commit, we try to step the sk_drops counter when tipc
is unable to queue a received message. Export sk_drops
using tipc socket diagnostics.

Acked-by: Jon Maloy <jon.maloy@...csson.com>
Acked-by: Ying Xue <ying.xue@...driver.com>
Signed-off-by: GhantaKrishnamurthy MohanKrishna <mohan.krishna.ghanta.krishnamurthy@...csson.com>
Signed-off-by: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan@...il.com>
---
 include/uapi/linux/tipc_netlink.h | 1 +
 net/tipc/socket.c                 | 9 +++++++--
 2 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/include/uapi/linux/tipc_netlink.h b/include/uapi/linux/tipc_netlink.h
index d7cec0480d70..d896ded51bcb 100644
--- a/include/uapi/linux/tipc_netlink.h
+++ b/include/uapi/linux/tipc_netlink.h
@@ -251,6 +251,7 @@ enum {
 	TIPC_NLA_SOCK_STAT_SENDQ,	/* u32 */
 	TIPC_NLA_SOCK_STAT_LINK_CONG,	/* flag */
 	TIPC_NLA_SOCK_STAT_CONN_CONG,	/* flag */
+	TIPC_NLA_SOCK_STAT_DROP,	/* u32 */
 
 	__TIPC_NLA_SOCK_STAT_MAX,
 	TIPC_NLA_SOCK_STAT_MAX = __TIPC_NLA_SOCK_STAT_MAX - 1
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 31fdd13d444e..a084c78408fb 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -2122,8 +2122,10 @@ static void tipc_sk_filter_rcv(struct sock *sk, struct sk_buff *skb,
 		    (!sk_conn && msg_connected(hdr)) ||
 		    (!grp && msg_in_group(hdr)))
 			err = TIPC_ERR_NO_PORT;
-		else if (sk_rmem_alloc_get(sk) + skb->truesize >= limit)
+		else if (sk_rmem_alloc_get(sk) + skb->truesize >= limit) {
+			atomic_inc(&sk->sk_drops);
 			err = TIPC_ERR_OVERLOAD;
+		}
 
 		if (unlikely(err)) {
 			tipc_skb_reject(net, err, skb, xmitq);
@@ -2202,6 +2204,7 @@ static void tipc_sk_enqueue(struct sk_buff_head *inputq, struct sock *sk,
 
 		/* Overload => reject message back to sender */
 		onode = tipc_own_addr(sock_net(sk));
+		atomic_inc(&sk->sk_drops);
 		if (tipc_msg_reverse(onode, &skb, TIPC_ERR_OVERLOAD))
 			__skb_queue_tail(xmitq, skb);
 		break;
@@ -3287,7 +3290,9 @@ int tipc_sk_fill_sock_diag(struct sk_buff *skb, struct tipc_sock *tsk,
 	if (nla_put_u32(skb, TIPC_NLA_SOCK_STAT_RCVQ,
 			skb_queue_len(&sk->sk_receive_queue)) ||
 	    nla_put_u32(skb, TIPC_NLA_SOCK_STAT_SENDQ,
-			skb_queue_len(&sk->sk_write_queue)))
+			skb_queue_len(&sk->sk_write_queue)) ||
+	    nla_put_u32(skb, TIPC_NLA_SOCK_STAT_DROP,
+			atomic_read(&sk->sk_drops)))
 		goto stat_msg_cancel;
 
 	if (tsk->cong_link_cnt &&
-- 
2.1.4