From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: [PATCH v4 net-next 19/19] inet: frags: get rid of nf_ct_frag6_skb_cb/NFCT_FRAG6_CB Date: Sat, 31 Mar 2018 12:59:00 -0700 Message-ID: <20180331195900.183604-20-edumazet@google.com> References: <20180331195900.183604-1-edumazet@google.com> Cc: netdev , Florian Westphal , Herbert Xu , Thomas Graf , Jesper Dangaard Brouer , Alexander Aring , Stefan Schmidt , Kirill Tkhai , Eric Dumazet , Eric Dumazet To: "David S . Miller" Return-path: Received: from mail-pf0-f196.google.com ([209.85.192.196]:34393 "EHLO mail-pf0-f196.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753074AbeCaUAA (ORCPT ); Sat, 31 Mar 2018 16:00:00 -0400 Received: by mail-pf0-f196.google.com with SMTP id q9so7481860pff.1 for ; Sat, 31 Mar 2018 13:00:00 -0700 (PDT) In-Reply-To: <20180331195900.183604-1-edumazet@google.com> Sender: netdev-owner@vger.kernel.org List-ID: nf_ct_frag6_queue() uses skb->cb[] to store the fragment offset, meaning that we could use two cache lines per skb when finding the insertion point, if for some reason inet6_skb_parm size is increased in the future. By using skb->ip_defrag_offset instead of skb->cb[] we pack all the fields in a single cache line, matching what we did for IPv4. Signed-off-by: Eric Dumazet --- net/ipv6/netfilter/nf_conntrack_reasm.c | 29 ++++++++++--------------- 1 file changed, 11 insertions(+), 18 deletions(-) diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index 603a395928593a071aff28105578d3bfdbf1b31e..3622aac343aea3cd0d8165d8681f42595735cd90 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -52,14 +52,6 @@ static const char nf_frags_cache_name[] = "nf-frags"; -struct nf_ct_frag6_skb_cb -{ - struct inet6_skb_parm h; - int offset; -}; - -#define NFCT_FRAG6_CB(skb) ((struct nf_ct_frag6_skb_cb *)((skb)->cb)) - static struct inet_frags nf_frags; #ifdef CONFIG_SYSCTL @@ -270,13 +262,13 @@ static int nf_ct_frag6_queue(struct frag_queue *fq, struct sk_buff *skb, * this fragment, right? */ prev = fq->q.fragments_tail; - if (!prev || NFCT_FRAG6_CB(prev)->offset < offset) { + if (!prev || prev->ip_defrag_offset < offset) { next = NULL; goto found; } prev = NULL; for (next = fq->q.fragments; next != NULL; next = next->next) { - if (NFCT_FRAG6_CB(next)->offset >= offset) + if (next->ip_defrag_offset >= offset) break; /* bingo! */ prev = next; } @@ -292,14 +284,19 @@ static int nf_ct_frag6_queue(struct frag_queue *fq, struct sk_buff *skb, /* Check for overlap with preceding fragment. */ if (prev && - (NFCT_FRAG6_CB(prev)->offset + prev->len) > offset) + (prev->ip_defrag_offset + prev->len) > offset) goto discard_fq; /* Look for overlap with succeeding segment. */ - if (next && NFCT_FRAG6_CB(next)->offset < end) + if (next && next->ip_defrag_offset < end) goto discard_fq; - NFCT_FRAG6_CB(skb)->offset = offset; + /* Note : skb->ip_defrag_offset and skb->dev share the same location */ + if (skb->dev) + fq->iif = skb->dev->ifindex; + /* Makes sure compiler wont do silly aliasing games */ + barrier(); + skb->ip_defrag_offset = offset; /* Insert this fragment in the chain of fragments. */ skb->next = next; @@ -310,10 +307,6 @@ static int nf_ct_frag6_queue(struct frag_queue *fq, struct sk_buff *skb, else fq->q.fragments = skb; - if (skb->dev) { - fq->iif = skb->dev->ifindex; - skb->dev = NULL; - } fq->q.stamp = skb->tstamp; fq->q.meat += skb->len; fq->ecn |= ecn; @@ -357,7 +350,7 @@ nf_ct_frag6_reasm(struct frag_queue *fq, struct sk_buff *prev, struct net_devic inet_frag_kill(&fq->q); WARN_ON(head == NULL); - WARN_ON(NFCT_FRAG6_CB(head)->offset != 0); + WARN_ON(head->ip_defrag_offset != 0); ecn = ip_frag_ecn_table[fq->ecn]; if (unlikely(ecn == 0xff)) -- 2.17.0.rc1.321.gba9d0f2565-goog