net: ipv4: fix RCU races on dst refcounts

commit c6cffba4ffa2 (ipv4: Fix input route performance regression.)
added various fatal races with dst refcounts.

crashes happen on tcp workloads if routes are added/deleted at the same
time.

The dst_free() calls from free_fib_info_rcu() are clearly racy.

We need instead regular dst refcounting (dst_release()) and make
sure dst_release() is aware of RCU grace periods :

Add DST_RCU_FREE flag so that dst_release() respects an RCU grace period
before dst destruction for cached dst

Introduce a new inet_sk_rx_dst_set() helper, using atomic_inc_not_zero()
to make sure we dont increase a zero refcount (On a dst currently
waiting an rcu grace period before destruction)

rt_cache_route() must take a reference on the new cached route, and
release it if was not able to install it.

With this patch, my machines survive various benchmarks.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 6 insertions, 6 deletions

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 2fbd992..7f91e5a 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1617,19 +1617,19 @@ int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)
 #endif
 
 	if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */
+		struct dst_entry *dst = sk->sk_rx_dst;
+
 		sock_rps_save_rxhash(sk, skb);
-		if (sk->sk_rx_dst) {
-			struct dst_entry *dst = sk->sk_rx_dst;
+		if (dst) {
 			if (inet_sk(sk)->rx_dst_ifindex != skb->skb_iif ||
 			    dst->ops->check(dst, 0) == NULL) {
 				dst_release(dst);
 				sk->sk_rx_dst = NULL;
 			}
 		}
-		if (unlikely(sk->sk_rx_dst == NULL)) {
-			sk->sk_rx_dst = dst_clone(skb_dst(skb));
-			inet_sk(sk)->rx_dst_ifindex = skb->skb_iif;
-		}
+		if (unlikely(sk->sk_rx_dst == NULL))
+			inet_sk_rx_dst_set(sk, skb);
+
 		if (tcp_rcv_established(sk, skb, tcp_hdr(skb), skb->len)) {
 			rsk = sk;
 			goto reset;