diff options
| author | Eric Dumazet <[email protected]> | 2010-04-28 21:35:48 +0000 |
|---|---|---|
| committer | David S. Miller <[email protected]> | 2010-04-28 21:35:48 +0000 |
| commit | 4b0b72f7dd617b13abd1b04c947e15873e011a24 (patch) | |
| tree | 16fc7bc990fa47cccb62bdb34cb23bd3c26b7a50 /net/ipv4/tcp_input.c | |
| parent | igb: Clean up left over prototype of igb_get_hw_dev_name() (diff) | |
| download | kernel-4b0b72f7dd617b13abd1b04c947e15873e011a24.tar.gz kernel-4b0b72f7dd617b13abd1b04c947e15873e011a24.zip | |
net: speedup udp receive path
Since commit 95766fff ([UDP]: Add memory accounting.),
each received packet needs one extra sock_lock()/sock_release() pair.
This added latency because of possible backlog handling. Then later,
ticket spinlocks added yet another latency source in case of DDOS.
This patch introduces lock_sock_bh() and unlock_sock_bh()
synchronization primitives, avoiding one atomic operation and backlog
processing.
skb_free_datagram_locked() uses them instead of full blown
lock_sock()/release_sock(). skb is orphaned inside locked section for
proper socket memory reclaim, and finally freed outside of it.
UDP receive path now take the socket spinlock only once.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Diffstat (limited to 'net/ipv4/tcp_input.c')
0 files changed, 0 insertions, 0 deletions