inetpeer: get rid of ip_id_count

[ Upstream commit 73f156a6e8c1074ac6327e0abd1169e95eb66463 ]

Ideally, we would need to generate IP ID using a per destination IP
generator.

linux kernels used inet_peer cache for this purpose, but this had a huge
cost on servers disabling MTU discovery.

1) each inet_peer struct consumes 192 bytes

2) inetpeer cache uses a binary tree of inet_peer structs,
   with a nominal size of ~66000 elements under load.

3) lookups in this tree are hitting a lot of cache lines, as tree depth
   is about 20.

4) If server deals with many tcp flows, we have a high probability of
   not finding the inet_peer, allocating a fresh one, inserting it in
   the tree with same initial ip_id_count, (cf secure_ip_id())

5) We garbage collect inet_peer aggressively.

IP ID generation do not have to be 'perfect'

Goal is trying to avoid duplicates in a short period of time,
so that reassembly units have a chance to complete reassembly of
fragments belonging to one message before receiving other fragments
with a recycled ID.

We simply use an array of generators, and a Jenkin hash using the dst IP
as a key.

ipv6_select_ident() is put back into net/ipv6/ip6_output.c where it
belongs (it is only used from this file)

secure_ip_id() and secure_ipv6_id() no longer are needed.

Rename ip_select_ident_more() to ip_select_ident_segs() to avoid
unnecessary decrement/increment of the number of segments.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>

diff --git a/drivers/net/ppp/pptp.c b/drivers/net/ppp/pptp.c
index 2b349d3..c974581 100644 (file)
--- a/drivers/net/ppp/pptp.c
+++ b/drivers/net/ppp/pptp.c
@@ -281,7 +281,7 @@ static int pptp_xmit(struct ppp_channel *chan, struct sk_buff *skb)
        nf_reset(skb);
 
        skb->ip_summed = CHECKSUM_NONE;
-       ip_select_ident(skb, &rt->dst, NULL);
+       ip_select_ident(skb, NULL);
        ip_send_check(iph);
 
        ip_local_out(skb);

diff --git a/include/net/inetpeer.h b/include/net/inetpeer.h
index ba0ddb0..6172a44 100644 (file)
--- a/include/net/inetpeer.h
+++ b/include/net/inetpeer.h
@@ -47,13 +47,12 @@ struct inet_peer {
        };
        /*
         * Once inet_peer is queued for deletion (refcnt == -1), following fields
-        * are not available: rid, ip_id_count, tcp_ts, tcp_ts_stamp
+        * are not available: rid, tcp_ts, tcp_ts_stamp
         * We can share memory with rcu_head to help keep inet_peer small.
         */
        union {
                struct {
                        atomic_t                        rid;            /* Frag reception counter */
-                       atomic_t                        ip_id_count;    /* IP ID for the next packet */
                        __u32                           tcp_ts;
                        __u32                           tcp_ts_stamp;
                };
@@ -103,7 +102,7 @@ extern bool inet_peer_xrlim_allow(struct inet_peer *peer, int timeout);
 extern void inetpeer_invalidate_tree(int family);
 
 /*
- * temporary check to make sure we dont access rid, ip_id_count, tcp_ts,
+ * temporary check to make sure we dont access rid, tcp_ts,
  * tcp_ts_stamp if no refcount is taken on inet_peer
  */
 static inline void inet_peer_refcheck(const struct inet_peer *p)
@@ -111,13 +110,4 @@ static inline void inet_peer_refcheck(const struct inet_peer *p)
        WARN_ON_ONCE(atomic_read(&p->refcnt) <= 0);
 }
 
-
-/* can be called with or without local BH being disabled */
-static inline int inet_getid(struct inet_peer *p, int more)
-{
-       more++;
-       inet_peer_refcheck(p);
-       return atomic_add_return(more, &p->ip_id_count) - more;
-}
-
 #endif /* _NET_INETPEER_H */

diff --cc include/net/ip.h
Simple merge

diff --cc include/net/ipip.h
Simple merge

diff --cc include/net/ipv6.h
Simple merge

diff --cc include/net/secure_seq.h
Simple merge

diff --cc net/core/secure_seq.c
Simple merge

diff --cc net/ipv4/igmp.c
Simple merge

diff --cc net/ipv4/inetpeer.c
Simple merge

diff --cc net/ipv4/ip_output.c
Simple merge

diff --cc net/ipv4/ipmr.c
Simple merge

diff --cc net/ipv4/raw.c
Simple merge

diff --cc net/ipv4/route.c
Simple merge

diff --cc net/ipv4/xfrm4_mode_tunnel.c
Simple merge

diff --cc net/ipv6/ip6_output.c
Simple merge

diff --cc net/netfilter/ipvs/ip_vs_xmit.c
Simple merge