From edf391ff17232f097d72441c9ad467bcb3b5db18 Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Mon, 27 Apr 2009 02:45:02 -0700 Subject: snmp: add missing counters for RFC 4293 The IP MIB (RFC 4293) defines stats for InOctets, OutOctets, InMcastOctets and OutMcastOctets: http://tools.ietf.org/html/rfc4293 But it seems we don't track those in any way that easy to separate from other protocols. This patch adds those missing counters to the stats file. Tested successfully by me With help from Eric Dumazet. Signed-off-by: Neil Horman Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'net/ipv4/ip_output.c') diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 3e7e910c7c0f..ea19c37ccc0c 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -181,10 +181,10 @@ static inline int ip_finish_output2(struct sk_buff *skb) struct net_device *dev = dst->dev; unsigned int hh_len = LL_RESERVED_SPACE(dev); - if (rt->rt_type == RTN_MULTICAST) - IP_INC_STATS(dev_net(dev), IPSTATS_MIB_OUTMCASTPKTS); - else if (rt->rt_type == RTN_BROADCAST) - IP_INC_STATS(dev_net(dev), IPSTATS_MIB_OUTBCASTPKTS); + if (rt->rt_type == RTN_MULTICAST) { + IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUTMCAST, skb->len); + } else if (rt->rt_type == RTN_BROADCAST) + IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUTBCAST, skb->len); /* Be paranoid, rather than too clever. */ if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { @@ -244,7 +244,7 @@ int ip_mc_output(struct sk_buff *skb) /* * If the indicated interface is up and running, send the packet. */ - IP_INC_STATS(dev_net(dev), IPSTATS_MIB_OUTREQUESTS); + IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUT, skb->len); skb->dev = dev; skb->protocol = htons(ETH_P_IP); @@ -298,7 +298,7 @@ int ip_output(struct sk_buff *skb) { struct net_device *dev = skb->dst->dev; - IP_INC_STATS(dev_net(dev), IPSTATS_MIB_OUTREQUESTS); + IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUT, skb->len); skb->dev = dev; skb->protocol = htons(ETH_P_IP); -- cgit v1.2.3 From 511c3f92ad5b6d9f8f6464be1b4f85f0422be91a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 2 Jun 2009 05:14:27 +0000 Subject: net: skb->rtable accessor Define skb_rtable(const struct sk_buff *skb) accessor to get rtable from skb Delete skb->rtable field Setting rtable is not allowed, just set dst instead as rtable is an alias. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'net/ipv4/ip_output.c') diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index ea19c37ccc0c..8d845ebfccaa 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -140,7 +140,7 @@ int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk, __be32 saddr, __be32 daddr, struct ip_options *opt) { struct inet_sock *inet = inet_sk(sk); - struct rtable *rt = skb->rtable; + struct rtable *rt = skb_rtable(skb); struct iphdr *iph; /* Build the IP header. */ @@ -238,7 +238,7 @@ static int ip_finish_output(struct sk_buff *skb) int ip_mc_output(struct sk_buff *skb) { struct sock *sk = skb->sk; - struct rtable *rt = skb->rtable; + struct rtable *rt = skb_rtable(skb); struct net_device *dev = rt->u.dst.dev; /* @@ -319,7 +319,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok) /* Skip all of this if the packet is already routed, * f.e. by something like SCTP. */ - rt = skb->rtable; + rt = skb_rtable(skb); if (rt != NULL) goto packet_routed; @@ -440,7 +440,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) unsigned int mtu, hlen, left, len, ll_rs, pad; int offset; __be16 not_last_frag; - struct rtable *rt = skb->rtable; + struct rtable *rt = skb_rtable(skb); int err = 0; dev = rt->u.dst.dev; @@ -1362,7 +1362,7 @@ void ip_send_reply(struct sock *sk, struct sk_buff *skb, struct ip_reply_arg *ar } replyopts; struct ipcm_cookie ipc; __be32 daddr; - struct rtable *rt = skb->rtable; + struct rtable *rt = skb_rtable(skb); if (ip_options_echo(&replyopts.opt, skb)) return; -- cgit v1.2.3 From adf30907d63893e4208dfe3f5c88ae12bc2f25d5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 2 Jun 2009 05:19:30 +0000 Subject: net: skb->dst accessors Define three accessors to get/set dst attached to a skb struct dst_entry *skb_dst(const struct sk_buff *skb) void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst) void skb_dst_drop(struct sk_buff *skb) This one should replace occurrences of : dst_release(skb->dst) skb->dst = NULL; Delete skb->dst field Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'net/ipv4/ip_output.c') diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 8d845ebfccaa..3d6167fb2d97 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -95,7 +95,7 @@ int __ip_local_out(struct sk_buff *skb) iph->tot_len = htons(skb->len); ip_send_check(iph); - return nf_hook(PF_INET, NF_INET_LOCAL_OUT, skb, NULL, skb->dst->dev, + return nf_hook(PF_INET, NF_INET_LOCAL_OUT, skb, NULL, skb_dst(skb)->dev, dst_output); } @@ -118,7 +118,7 @@ static int ip_dev_loopback_xmit(struct sk_buff *newskb) __skb_pull(newskb, skb_network_offset(newskb)); newskb->pkt_type = PACKET_LOOPBACK; newskb->ip_summed = CHECKSUM_UNNECESSARY; - WARN_ON(!newskb->dst); + WARN_ON(!skb_dst(newskb)); netif_rx(newskb); return 0; } @@ -176,7 +176,7 @@ EXPORT_SYMBOL_GPL(ip_build_and_send_pkt); static inline int ip_finish_output2(struct sk_buff *skb) { - struct dst_entry *dst = skb->dst; + struct dst_entry *dst = skb_dst(skb); struct rtable *rt = (struct rtable *)dst; struct net_device *dev = dst->dev; unsigned int hh_len = LL_RESERVED_SPACE(dev); @@ -217,14 +217,14 @@ static inline int ip_skb_dst_mtu(struct sk_buff *skb) struct inet_sock *inet = skb->sk ? inet_sk(skb->sk) : NULL; return (inet && inet->pmtudisc == IP_PMTUDISC_PROBE) ? - skb->dst->dev->mtu : dst_mtu(skb->dst); + skb_dst(skb)->dev->mtu : dst_mtu(skb_dst(skb)); } static int ip_finish_output(struct sk_buff *skb) { #if defined(CONFIG_NETFILTER) && defined(CONFIG_XFRM) /* Policy lookup after SNAT yielded a new policy */ - if (skb->dst->xfrm != NULL) { + if (skb_dst(skb)->xfrm != NULL) { IPCB(skb)->flags |= IPSKB_REROUTED; return dst_output(skb); } @@ -296,7 +296,7 @@ int ip_mc_output(struct sk_buff *skb) int ip_output(struct sk_buff *skb) { - struct net_device *dev = skb->dst->dev; + struct net_device *dev = skb_dst(skb)->dev; IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUT, skb->len); @@ -355,7 +355,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok) } sk_setup_caps(sk, &rt->u.dst); } - skb->dst = dst_clone(&rt->u.dst); + skb_dst_set(skb, dst_clone(&rt->u.dst)); packet_routed: if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) @@ -401,8 +401,8 @@ static void ip_copy_metadata(struct sk_buff *to, struct sk_buff *from) to->pkt_type = from->pkt_type; to->priority = from->priority; to->protocol = from->protocol; - dst_release(to->dst); - to->dst = dst_clone(from->dst); + skb_dst_drop(to); + skb_dst_set(to, dst_clone(skb_dst(from))); to->dev = from->dev; to->mark = from->mark; @@ -1294,7 +1294,7 @@ int ip_push_pending_frames(struct sock *sk) * on dst refcount */ inet->cork.dst = NULL; - skb->dst = &rt->u.dst; + skb_dst_set(skb, &rt->u.dst); if (iph->protocol == IPPROTO_ICMP) icmp_out_count(net, ((struct icmphdr *) -- cgit v1.2.3 From d7fcf1a5cae2c970e9afe7192fe0c13d931247e0 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 9 Jun 2009 00:19:37 -0700 Subject: ipv4: Use frag list abstraction interfaces. Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'net/ipv4/ip_output.c') diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 3d6167fb2d97..9248d2807ba6 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -474,7 +474,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) * LATER: this step can be merged to real generation of fragments, * we can switch to copy when see the first bad fragment. */ - if (skb_shinfo(skb)->frag_list) { + if (skb_has_frags(skb)) { struct sk_buff *frag; int first_len = skb_pagelen(skb); int truesizes = 0; @@ -485,7 +485,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) skb_cloned(skb)) goto slow_path; - for (frag = skb_shinfo(skb)->frag_list; frag; frag = frag->next) { + skb_walk_frags(skb, frag) { /* Correct geometry. */ if (frag->len > mtu || ((frag->len & 7) && frag->next) || @@ -510,7 +510,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) err = 0; offset = 0; frag = skb_shinfo(skb)->frag_list; - skb_shinfo(skb)->frag_list = NULL; + skb_frag_list_init(skb); skb->data_len = first_len - skb_headlen(skb); skb->truesize -= truesizes; skb->len = first_len; -- cgit v1.2.3 From 2b85a34e911bf483c27cfdd124aeb1605145dc80 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 11 Jun 2009 02:55:43 -0700 Subject: net: No more expensive sock_hold()/sock_put() on each tx One of the problem with sock memory accounting is it uses a pair of sock_hold()/sock_put() for each transmitted packet. This slows down bidirectional flows because the receive path also needs to take a refcount on socket and might use a different cpu than transmit path or transmit completion path. So these two atomic operations also trigger cache line bounces. We can see this in tx or tx/rx workloads (media gateways for example), where sock_wfree() can be in top five functions in profiles. We use this sock_hold()/sock_put() so that sock freeing is delayed until all tx packets are completed. As we also update sk_wmem_alloc, we could offset sk_wmem_alloc by one unit at init time, until sk_free() is called. Once sk_free() is called, we atomic_dec_and_test(sk_wmem_alloc) to decrement initial offset and atomicaly check if any packets are in flight. skb_set_owner_w() doesnt call sock_hold() anymore sock_wfree() doesnt call sock_put() anymore, but check if sk_wmem_alloc reached 0 to perform the final freeing. Drawback is that a skb->truesize error could lead to unfreeable sockets, or even worse, prematurely calling __sk_free() on a live socket. Nice speedups on SMP. tbench for example, going from 2691 MB/s to 2711 MB/s on my 8 cpu dev machine, even if tbench was not really hitting sk_refcnt contention point. 5 % speedup on a UDP transmit workload (depends on number of flows), lowering TX completion cpu usage. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/ip_output.c | 1 - 1 file changed, 1 deletion(-) (limited to 'net/ipv4/ip_output.c') diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 9248d2807ba6..247026282669 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -498,7 +498,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *)) BUG_ON(frag->sk); if (skb->sk) { - sock_hold(skb->sk); frag->sk = skb->sk; frag->destructor = sock_wfree; truesizes += frag->truesize; -- cgit v1.2.3