diff options
| author | Jakub Kicinski <kuba@kernel.org> | 2025-04-08 18:19:45 -0700 |
|---|---|---|
| committer | Jakub Kicinski <kuba@kernel.org> | 2025-04-08 18:19:45 -0700 |
| commit | 6a0f72c5e208c5980fe4bb7d0c2ecc31714207c4 (patch) | |
| tree | 42374557872d888984158a19d5abbb2f408129d3 /include/net/udp.h | |
| parent | 5ac40e6b5b0bac3f72020a48d01ada23e8450d0c (diff) | |
| parent | 5d7f5b2f6b935517ee5fd8058dc32342a5cba3e1 (diff) | |
| download | linux-next-6a0f72c5e208c5980fe4bb7d0c2ecc31714207c4.tar.gz linux-next-6a0f72c5e208c5980fe4bb7d0c2ecc31714207c4.zip | |
Merge branch 'udp_tunnel-gro-optimizations'
Paolo Abeni says:
====================
udp_tunnel: GRO optimizations
The UDP tunnel GRO stage is source of measurable overhead for workload
based on UDP-encapsulated traffic: each incoming packets requires a full
UDP socket lookup and an indirect call.
In the most common setups a single UDP tunnel device is used. In such
case we can optimize both the lookup and the indirect call.
Patch 1 tracks per netns the active UDP tunnels and replaces the socket
lookup with a single destination port comparison when possible.
Patch 2 tracks the different types of UDP tunnels and replaces the
indirect call with a static one when there is a single UDP tunnel type
active.
I measure ~10% performance improvement in TCP over UDP tunnel stream
tests on top of this series.
v4: https://lore.kernel.org/cover.1741718157.git.pabeni@redhat.com
v3: https://lore.kernel.org/cover.1741632298.git.pabeni@redhat.com
v2: https://lore.kernel.org/cover.1741338765.git.pabeni@redhat.com
v1: https://lore.kernel.org/cover.1741275846.git.pabeni@redhat.com
====================
Link: https://patch.msgid.link/cover.1744040675.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Diffstat (limited to 'include/net/udp.h')
| -rw-r--r-- | include/net/udp.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/include/net/udp.h b/include/net/udp.h index 6e89520e100d..a772510b2aa5 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -290,6 +290,7 @@ static inline void udp_lib_init_sock(struct sock *sk) struct udp_sock *up = udp_sk(sk); skb_queue_head_init(&up->reader_queue); + INIT_HLIST_NODE(&up->tunnel_list); up->forward_threshold = sk->sk_rcvbuf >> 2; set_bit(SOCK_CUSTOM_SOCKOPT, &sk->sk_socket->flags); } |