From 87fb4b7b533073eeeaed0b6bf7c2328995f6c075 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 13 Oct 2011 07:28:54 +0000 Subject: net: more accurate skb truesize skb truesize currently accounts for sk_buff struct and part of skb head. kmalloc() roundings are also ignored. Considering that skb_shared_info is larger than sk_buff, its time to take it into account for better memory accounting. This patch introduces SKB_TRUESIZE(X) macro to centralize various assumptions into a single place. At skb alloc phase, we put skb_shared_info struct at the exact end of skb head, to allow a better use of memory (lowering number of reallocations), since kmalloc() gives us power-of-two memory blocks. Unless SLUB/SLUB debug is active, both skb->head and skb_shared_info are aligned to cache lines, as before. Note: This patch might trigger performance regressions because of misconfigured protocol stacks, hitting per socket or global memory limits that were previously not reached. But its a necessary step for a more accurate memory accounting. Signed-off-by: Eric Dumazet CC: Andi Kleen CC: Ben Hutchings Signed-off-by: David S. Miller --- net/ipv4/icmp.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'net/ipv4/icmp.c') diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 23ef31baa1af..ab188ae12fd9 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -1152,10 +1152,9 @@ static int __net_init icmp_sk_init(struct net *net) net->ipv4.icmp_sk[i] = sk; /* Enough space for 2 64K ICMP packets, including - * sk_buff struct overhead. + * sk_buff/skb_shared_info struct overhead. */ - sk->sk_sndbuf = - (2 * ((64 * 1024) + sizeof(struct sk_buff))); + sk->sk_sndbuf = 2 * SKB_TRUESIZE(64 * 1024); /* * Speedup sock_wfree() -- cgit v1.2.3