mbox series

[net-next,v4,0/2] tcp: make trace of reset logic complete

Message ID 20240401073605.37335-1-kerneljasonxing@gmail.com (mailing list archive)
Headers show
Series tcp: make trace of reset logic complete | expand

Message

Jason Xing April 1, 2024, 7:36 a.m. UTC
From: Jason Xing <kernelxing@tencent.com>

Before this, we miss some cases where the TCP layer could send RST but
we cannot trace it. So I decided to complete it :)

v4
Link: https://lore.kernel.org/all/20240329034243.7929-1-kerneljasonxing@gmail.com/
1. rebased against latest net-next
2. remove {} and add skb test statement (Eric)
3. drop v3 patch [3/3] temporarily because 1) location is not that useful
since we can use perf or something else to trace, 2) Eric said we could
use drop_reason to show why we have to RST, which is good, but this seems
not work well for those ->send_reset() logic. I need more time to
investigate this part.

v3
1. fix a format problem in patch [3/3]

v2
1. fix spelling mistakes

Jason Xing (2):
  trace: adjust TP_STORE_ADDR_PORTS_SKB() parameters
  trace: tcp: fully support trace_tcp_send_reset

 include/trace/events/net_probe_common.h | 20 ++++++------
 include/trace/events/tcp.h              | 42 +++++++++++++++++++++++--
 include/trace/events/udp.h              |  2 +-
 net/ipv4/tcp_ipv4.c                     |  7 ++---
 net/ipv6/tcp_ipv6.c                     |  3 +-
 5 files changed, 56 insertions(+), 18 deletions(-)

Comments

patchwork-bot+netdevbpf@kernel.org April 4, 2024, 2:30 a.m. UTC | #1
Hello:

This series was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Mon,  1 Apr 2024 15:36:03 +0800 you wrote:
> From: Jason Xing <kernelxing@tencent.com>
> 
> Before this, we miss some cases where the TCP layer could send RST but
> we cannot trace it. So I decided to complete it :)
> 
> v4
> Link: https://lore.kernel.org/all/20240329034243.7929-1-kerneljasonxing@gmail.com/
> 1. rebased against latest net-next
> 2. remove {} and add skb test statement (Eric)
> 3. drop v3 patch [3/3] temporarily because 1) location is not that useful
> since we can use perf or something else to trace, 2) Eric said we could
> use drop_reason to show why we have to RST, which is good, but this seems
> not work well for those ->send_reset() logic. I need more time to
> investigate this part.
> 
> [...]

Here is the summary with links:
  - [net-next,v4,1/2] trace: adjust TP_STORE_ADDR_PORTS_SKB() parameters
    https://git.kernel.org/netdev/net-next/c/9807080e2170
  - [net-next,v4,2/2] trace: tcp: fully support trace_tcp_send_reset
    https://git.kernel.org/netdev/net-next/c/19822a980e19

You are awesome, thank you!