gro: Optimise TCP packet reception

gro: Optimise TCP packet reception

As this function can be called more than half a million times for
10GbE, it's important to optimise it as much as we can.

This patch uses bit ops to logical ops, as well as open coding
memcmp to exploit alignment properties.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Herbert Xu 2009-02-08 18:00:40 +00:00 committed by David S. Miller
parent a5ad24be72
commit aa6320d336
1 changed files with 8 additions and 7 deletions

View File

@ -2478,9 +2478,9 @@ struct sk_buff **tcp_gro_receive(struct sk_buff **head, struct sk_buff *skb)
struct tcphdr *th2;
unsigned int thlen;
unsigned int flags;
unsigned int total;
unsigned int mss = 1;
int flush = 1;
int i;
th = skb_gro_header(skb, sizeof(*th));
if (unlikely(!th))
@ -2504,7 +2504,7 @@ struct sk_buff **tcp_gro_receive(struct sk_buff **head, struct sk_buff *skb)
th2 = tcp_hdr(p);
if (th->source != th2->source || th->dest != th2->dest) {
if ((th->source ^ th2->source) | (th->dest ^ th2->dest)) {
NAPI_GRO_CB(p)->same_flow = 0;
continue;
}
@ -2519,14 +2519,15 @@ found:
flush |= flags & TCP_FLAG_CWR;
flush |= (flags ^ tcp_flag_word(th2)) &
~(TCP_FLAG_CWR | TCP_FLAG_FIN | TCP_FLAG_PSH);
flush |= th->ack_seq != th2->ack_seq || th->window != th2->window;
flush |= memcmp(th + 1, th2 + 1, thlen - sizeof(*th));
flush |= (th->ack_seq ^ th2->ack_seq) | (th->window ^ th2->window);
for (i = sizeof(*th); !flush && i < thlen; i += 4)
flush |= *(u32 *)((u8 *)th + i) ^
*(u32 *)((u8 *)th2 + i);
total = skb_gro_len(p);
mss = skb_shinfo(p)->gso_size;
flush |= skb_gro_len(skb) > mss || !skb_gro_len(skb);
flush |= ntohl(th2->seq) + total != ntohl(th->seq);
flush |= (skb_gro_len(skb) > mss) | !skb_gro_len(skb);
flush |= (ntohl(th2->seq) + skb_gro_len(p)) ^ ntohl(th->seq);
if (flush || skb_gro_receive(head, skb)) {
mss = 1;