From d826eb14ecef3574b6b3be55e5f4329f4a76fbf3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 9 Nov 2011 07:24:35 +0000 Subject: ipv4: PKTINFO doesnt need dst reference MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Le lundi 07 novembre 2011 à 15:33 +0100, Eric Dumazet a écrit : > At least, in recent kernels we dont change dst->refcnt in forwarding > patch (usinf NOREF skb->dst) > > One particular point is the atomic_inc(dst->refcnt) we have to perform > when queuing an UDP packet if socket asked PKTINFO stuff (for example a > typical DNS server has to setup this option) > > I have one patch somewhere that stores the information in skb->cb[] and > avoid the atomic_{inc|dec}(dst->refcnt). > OK I found it, I did some extra tests and believe its ready. [PATCH net-next] ipv4: IP_PKTINFO doesnt need dst reference When a socket uses IP_PKTINFO notifications, we currently force a dst reference for each received skb. Reader has to access dst to get needed information (rt_iif & rt_spec_dst) and must release dst reference. We also forced a dst reference if skb was put in socket backlog, even without IP_PKTINFO handling. This happens under stress/load. We can instead store the needed information in skb->cb[], so that only softirq handler really access dst, improving cache hit ratios. This removes two atomic operations per packet, and false sharing as well. On a benchmark using a mono threaded receiver (doing only recvmsg() calls), I can reach 720.000 pps instead of 570.000 pps. IP_PKTINFO is typically used by DNS servers, and any multihomed aware UDP application. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/raw.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/ipv4/raw.c') diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 007e2eb769d3..7a8410d1b4b1 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -292,7 +292,8 @@ static int raw_rcv_skb(struct sock * sk, struct sk_buff * skb) { /* Charge it to the socket. */ - if (ip_queue_rcv_skb(sk, skb) < 0) { + ipv4_pktinfo_prepare(skb); + if (sock_queue_rcv_skb(sk, skb) < 0) { kfree_skb(skb); return NET_RX_DROP; } -- cgit v1.2.3 From 660882432909dbe611f1792eda158188065cb9f1 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Fri, 18 Nov 2011 02:20:04 +0000 Subject: ipv4: Remove all uses of LL_ALLOCATED_SPACE ipv4: Remove all uses of LL_ALLOCATED_SPACE The macro LL_ALLOCATED_SPACE was ill-conceived. It applies the alignment to the sum of needed_headroom and needed_tailroom. As the amount that is then reserved for head room is needed_headroom with alignment, this means that the tail room left may be too small. This patch replaces all uses of LL_ALLOCATED_SPACE in net/ipv4 with the macro LL_RESERVED_SPACE and direct reference to needed_tailroom. This also fixes the problem with needed_headroom changing between allocating the skb and reserving the head room. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/ipv4/raw.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'net/ipv4/raw.c') diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 7a8410d1b4b1..3ccda5ae8a27 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -328,6 +328,7 @@ static int raw_send_hdrinc(struct sock *sk, struct flowi4 *fl4, unsigned int iphlen; int err; struct rtable *rt = *rtp; + int hlen, tlen; if (length > rt->dst.dev->mtu) { ip_local_error(sk, EMSGSIZE, fl4->daddr, inet->inet_dport, @@ -337,12 +338,14 @@ static int raw_send_hdrinc(struct sock *sk, struct flowi4 *fl4, if (flags&MSG_PROBE) goto out; + hlen = LL_RESERVED_SPACE(rt->dst.dev); + tlen = rt->dst.dev->needed_tailroom; skb = sock_alloc_send_skb(sk, - length + LL_ALLOCATED_SPACE(rt->dst.dev) + 15, + length + hlen + tlen + 15, flags & MSG_DONTWAIT, &err); if (skb == NULL) goto error; - skb_reserve(skb, LL_RESERVED_SPACE(rt->dst.dev)); + skb_reserve(skb, hlen); skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; -- cgit v1.2.3