[net-next,0/5] tcp: allow to reduce max RTO

Message

Eric Dumazet Feb. 7, 2025, 3:28 p.m. UTC

This is a followup of a discussion started 6 months ago
by Jason Xing.

Some applications want to lower the time between each
retransmit attempts.

TCP_KEEPINTVL and TCP_KEEPCNT socket options don't
work around the issue.

This series adds:

- a new TCP level socket option (TCP_RTO_MAX_MS)
- a new sysctl (/proc/sys/net/ipv4/tcp_rto_max_ms)

Admins and/or applications can now change the max rto value
at their own risk.

Link: https://lore.kernel.org/netdev/[e-mail]/T/

Eric Dumazet (5):
  tcp: remove tcp_reset_xmit_timer() @max_when argument
  tcp: add a @pace_delay parameter to tcp_reset_xmit_timer()
  tcp: use tcp_reset_xmit_timer()
  tcp: add the ability to control max RTO
  tcp: add tcp_rto_max_ms sysctl

 Documentation/networking/ip-sysctl.rst        | 13 ++++++++
 .../net_cachelines/inet_connection_sock.rst   |  1 +
 .../net_cachelines/netns_ipv4_sysctl.rst      |  1 +
 include/net/inet_connection_sock.h            |  1 +
 include/net/netns/ipv4.h                      |  1 +
 include/net/tcp.h                             | 23 +++++++++-----
 include/uapi/linux/tcp.h                      |  1 +
 net/ipv4/sysctl_net_ipv4.c                    | 10 ++++++
 net/ipv4/tcp.c                                | 14 ++++++++-
 net/ipv4/tcp_fastopen.c                       |  4 +--
 net/ipv4/tcp_input.c                          | 19 +++++-------
 net/ipv4/tcp_ipv4.c                           |  6 ++--
 net/ipv4/tcp_output.c                         | 17 +++++-----
 net/ipv4/tcp_timer.c                          | 31 ++++++++++---------
 14 files changed, 93 insertions(+), 49 deletions(-)

Comments

[e-mail] Feb. 11, 2025, 12:30 p.m. UTC | #1

Hello:

This series was applied to netdev/net-next.git (main)
by Paolo Abeni <[e-mail]>:

On Fri,  7 Feb 2025 15:28:25 +0000 you wrote:
> This is a followup of a discussion started 6 months ago
> by Jason Xing.
> 
> Some applications want to lower the time between each
> retransmit attempts.
> 
> TCP_KEEPINTVL and TCP_KEEPCNT socket options don't
> work around the issue.
> 
> [...]

Here is the summary with links:
  - [net-next,1/5] tcp: remove tcp_reset_xmit_timer() @max_when argument
    https://git.kernel.org/netdev/net-next/c/0fed463777b8
  - [net-next,2/5] tcp: add a @pace_delay parameter to tcp_reset_xmit_timer()
    https://git.kernel.org/netdev/net-next/c/7baa030155e8
  - [net-next,3/5] tcp: use tcp_reset_xmit_timer()
    https://git.kernel.org/netdev/net-next/c/48b69b4c7e5d
  - [net-next,4/5] tcp: add the ability to control max RTO
    https://git.kernel.org/netdev/net-next/c/54a378f43425
  - [net-next,5/5] tcp: add tcp_rto_max_ms sysctl
    https://git.kernel.org/netdev/net-next/c/1280c26228bd

You are awesome, thank you!