diff options
| author | Kui-Feng Lee <kuifeng@meta.com> | 2023-03-23 06:23:59 +0300 |
|---|---|---|
| committer | Martin KaFai Lau <martin.lau@kernel.org> | 2023-03-23 08:53:00 +0300 |
| commit | 8fb1a76a0f35c45a424c9eb84b0f97ffd51e6052 (patch) | |
| tree | fbe167ee5d52342b50005ab130fdc88d7e4c02e8 /include/net/tcp.h | |
| parent | b671c2067a04c0668df174ff5dfdb573d1f9b074 (diff) | |
| download | linux-8fb1a76a0f35c45a424c9eb84b0f97ffd51e6052.tar.xz | |
net: Update an existing TCP congestion control algorithm.
This feature lets you immediately transition to another congestion
control algorithm or implementation with the same name. Once a name
is updated, new connections will apply this new algorithm.
The purpose is to update a customized algorithm implemented in BPF
struct_ops with a new version on the flight. The following is an
example of using the userspace API implemented in later BPF patches.
link = bpf_map__attach_struct_ops(skel->maps.ca_update_1);
.......
err = bpf_link__update_map(link, skel->maps.ca_update_2);
We first load and register an algorithm implemented in BPF struct_ops,
then swap it out with a new one using the same name. After that, newly
created connections will apply the updated algorithm, while older ones
retain the previous version already applied.
This patch also takes this chance to refactor the ca validation into
the new tcp_validate_congestion_control() function.
Cc: netdev@vger.kernel.org, Eric Dumazet <edumazet@google.com>
Signed-off-by: Kui-Feng Lee <kuifeng@meta.com>
Link: https://lore.kernel.org/r/20230323032405.3735486-3-kuifeng@meta.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Diffstat (limited to 'include/net/tcp.h')
| -rw-r--r-- | include/net/tcp.h | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/include/net/tcp.h b/include/net/tcp.h index db9f828e9d1e..2abb755e6a3a 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1117,6 +1117,9 @@ struct tcp_congestion_ops { int tcp_register_congestion_control(struct tcp_congestion_ops *type); void tcp_unregister_congestion_control(struct tcp_congestion_ops *type); +int tcp_update_congestion_control(struct tcp_congestion_ops *type, + struct tcp_congestion_ops *old_type); +int tcp_validate_congestion_control(struct tcp_congestion_ops *ca); void tcp_assign_congestion_control(struct sock *sk); void tcp_init_congestion_control(struct sock *sk); |