tipc: reduce risk of user starvation during link congestion

The socket code currently handles link congestion by either blocking
and trying to send again when the congestion has abated, or just
returning to the user with -EAGAIN and let him re-try later.

This mechanism is prone to starvation, because the wakeup algorithm is
non-atomic. During the time the link issues a wakeup signal, until the
socket wakes up and re-attempts sending, other senders may have come
in between and occupied the free buffer space in the link. This in turn
may lead to a socket having to make many send attempts before it is
successful. In extremely loaded systems we have observed latency times
of several seconds before a low-priority socket is able to send out a
message.

In this commit, we simplify this mechanism and reduce the risk of the
described scenario happening. When a message is attempted sent via a
congested link, we now let it be added to the link's backlog queue
anyway, thus permitting an oversubscription of one message per source
socket. We still create a wakeup item and return an error code, hence
instructing the sender to block or stop sending. Only when enough space
has been freed up in the link's backlog queue do we issue a wakeup event
that allows the sender to continue with the next message, if any.

The fact that a socket now can consider a message sent even when the
link returns a congestion code means that the sending socket code can
be simplified. Also, since this is a good opportunity to get rid of the
obsolete 'mtu change' condition in the three socket send functions, we
now choose to refactor those functions completely.

Signed-off-by: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 5 insertions, 10 deletions

diff --git a/net/tipc/node.c b/net/tipc/node.c
index 9d2f4c2b08ab..2883f6a0ed98 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -1167,7 +1167,7 @@ msg_full:
  * @list: chain of buffers containing message
  * @dnode: address of destination node
  * @selector: a number used for deterministic link selection
- * Consumes the buffer chain, except when returning -ELINKCONG
+ * Consumes the buffer chain.
  * Returns 0 if success, otherwise: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE,-ENOBUF
  */
 int tipc_node_xmit(struct net *net, struct sk_buff_head *list,
@@ -1206,10 +1206,10 @@ int tipc_node_xmit(struct net *net, struct sk_buff_head *list,
         spin_unlock_bh(&le->lock);
         tipc_node_read_unlock(n);
 
-        if (likely(rc == 0))
-                tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr);
-        else if (rc == -ENOBUFS)
+        if (unlikely(rc == -ENOBUFS))
                 tipc_node_link_down(n, bearer_id, false);
+        else
+                tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr);
 
         tipc_node_put(n);
 
@@ -1221,20 +1221,15 @@ int tipc_node_xmit(struct net *net, struct sk_buff_head *list,
  * messages, which will not be rejected
  * The only exception is datagram messages rerouted after secondary
  * lookup, which are rare and safe to dispose of anyway.
- * TODO: Return real return value, and let callers use
- * tipc_wait_for_sendpkt() where applicable
  */
 int tipc_node_xmit_skb(struct net *net, struct sk_buff *skb, u32 dnode,
                        u32 selector)
 {
         struct sk_buff_head head;
-        int rc;
 
         skb_queue_head_init(&head);
         __skb_queue_tail(&head, skb);
-        rc = tipc_node_xmit(net, &head, dnode, selector);
-        if (rc == -ELINKCONG)
-                kfree_skb(skb);
+        tipc_node_xmit(net, &head, dnode, selector);
         return 0;
 }