udp: be less conservative with sock rmem accounting

Before commit 850cbaddb52d ("udp: use it's own memory accounting
schema"), the udp protocol allowed sk_rmem_alloc to grow beyond
the rcvbuf by the whole current packet's truesize. After said commit
we allow sk_rmem_alloc to exceed the rcvbuf only if the receive queue
is empty. As reported by Jesper this cause a performance regression
for some (small) values of rcvbuf.

This commit is intended to fix the regression restoring the old
handling of the rcvbuf limit.

Reported-by: Jesper Dangaard Brouer <brouer@redhat.com>
Fixes: 850cbaddb52d ("udp: use it's own memory accounting schema")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index e1d0bf8..16d88ba 100644 (file)
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1205,14 +1205,14 @@ int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb)
         * queue is full; always allow at least a packet
         */
        rmem = atomic_read(&sk->sk_rmem_alloc);
-       if (rmem && (rmem + size > sk->sk_rcvbuf))
+       if (rmem > sk->sk_rcvbuf)
                goto drop;
 
        /* we drop only if the receive buf is full and the receive
         * queue contains some other skb
         */
        rmem = atomic_add_return(size, &sk->sk_rmem_alloc);
-       if ((rmem > sk->sk_rcvbuf) && (rmem > size))
+       if (rmem > (size + sk->sk_rcvbuf))
                goto uncharge_drop;
 
        spin_lock(&list->lock);