diff options
| author | David S. Miller <davem@davemloft.net> | 2005-08-18 01:57:30 +0400 |
|---|---|---|
| committer | David S. Miller <davem@sunset.davemloft.net> | 2005-08-30 03:01:54 +0400 |
| commit | d179cd12928443f3ec29cfbc3567439644bd0afc (patch) | |
| tree | 0bfc57e73f0bf9f7bb9d5c8ce7d3d5afe550f94e /net/ipv4 | |
| parent | e92ae93a8aa66aea12935420cb22d4df1c18d023 (diff) | |
| download | linux-d179cd12928443f3ec29cfbc3567439644bd0afc.tar.xz | |
[NET]: Implement SKB fast cloning.
Protocols that make extensive use of SKB cloning,
for example TCP, eat at least 2 allocations per
packet sent as a result.
To cut the kmalloc() count in half, we implement
a pre-allocation scheme wherein we allocate
2 sk_buff objects in advance, then use a simple
reference count to free up the memory at the
correct time.
Based upon an initial patch by Thomas Graf and
suggestions from Herbert Xu.
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4')
| -rw-r--r-- | net/ipv4/tcp_output.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 8d92ab562aed..75b68116682a 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -1582,7 +1582,7 @@ void tcp_send_fin(struct sock *sk) } else { /* Socket is locked, keep trying until memory is available. */ for (;;) { - skb = alloc_skb(MAX_TCP_HEADER, GFP_KERNEL); + skb = alloc_skb_fclone(MAX_TCP_HEADER, GFP_KERNEL); if (skb) break; yield(); @@ -1804,7 +1804,7 @@ int tcp_connect(struct sock *sk) tcp_connect_init(sk); - buff = alloc_skb(MAX_TCP_HEADER + 15, sk->sk_allocation); + buff = alloc_skb_fclone(MAX_TCP_HEADER + 15, sk->sk_allocation); if (unlikely(buff == NULL)) return -ENOBUFS; |