Here are various fixes:

- Patch 1: Better handling SKB extension allocation failures. A fix for
  v5.7.

- Patches 2, 3: Avoid resetting MPTCP limits when flushing MPTCP
  endpoints. With a validation in the selftests. Fixes for v5.7.

- Patches 4, 5, 6: Disallow '0' as ADD_ADDR retransmission timeout.
  With a preparation patch, and a validation in the selftests. Fixes for
  v5.11.

- Patches 8, 9: Fix C23 extension warnings in the selftests, spotted by
  GCC. Fixes for v6.16.

Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
Christoph Paasch (1):
      mptcp: drop skb if MPTCP skb extension allocation fails

Geliang Tang (3):
      mptcp: remove duplicate sk_reset_timer call
      mptcp: disable add_addr retransmission when timeout is 0
      selftests: mptcp: disable add_addr retrans in endpoint_tests

Matthieu Baerts (NGI0) (4):
      mptcp: pm: kernel: flush: do not reset ADD_ADDR limit
      selftests: mptcp: pm: check flush doesn't reset limits
      selftests: mptcp: connect: fix C23 extension warning
      selftests: mptcp: sockopt: fix C23 extension warning

 Documentation/networking/mptcp-sysctl.rst         |  2 ++
 net/mptcp/options.c                               |  6 ++++--
 net/mptcp/pm.c                                    | 18 ++++++++++++------
 net/mptcp/pm_kernel.c                             |  1 -
 tools/testing/selftests/net/mptcp/mptcp_connect.c |  5 +++--
 tools/testing/selftests/net/mptcp/mptcp_inq.c     |  5 +++--
 tools/testing/selftests/net/mptcp/mptcp_join.sh   |  1 +
 tools/testing/selftests/net/mptcp/mptcp_sockopt.c |  5 +++--
 tools/testing/selftests/net/mptcp/pm_netlink.sh   |  1 +
 9 files changed, 29 insertions(+), 15 deletions(-)
---
base-commit: 065c31f2c6915b38f45b1c817b31f41f62eaa774
change-id: 20250815-net-mptcp-misc-fixes-6-17-rc2-d18b2437e8d4

Best regards,
-- 
Matthieu Baerts (NGI0) <matttbe@kernel.org>

On Fri, 15 Aug 2025 19:28:18 +0200 Matthieu Baerts (NGI0) wrote:
> Here are various fixes:
> 
> - Patch 1: Better handling SKB extension allocation failures. A fix for
>   v5.7.
> 
> - Patches 2, 3: Avoid resetting MPTCP limits when flushing MPTCP
>   endpoints. With a validation in the selftests. Fixes for v5.7.
> 
> - Patches 4, 5, 6: Disallow '0' as ADD_ADDR retransmission timeout.
>   With a preparation patch, and a validation in the selftests. Fixes for
>   v5.11.
> 
> - Patches 8, 9: Fix C23 extension warnings in the selftests, spotted by
>   GCC. Fixes for v6.16.

userspace_pm.sh which hasn't flaked in 1000 runs has flaked last night,
with this series applied:
https://netdev-3.bots.linux.dev/vmksft-mptcp/results/255941/8-userspace-pm-sh/
Looks unrelated but also quite strange?

Hi Jakub,

On 16/08/2025 20:27, Jakub Kicinski wrote:
> On Fri, 15 Aug 2025 19:28:18 +0200 Matthieu Baerts (NGI0) wrote:
>> Here are various fixes:
>>
>> - Patch 1: Better handling SKB extension allocation failures. A fix for
>>   v5.7.
>>
>> - Patches 2, 3: Avoid resetting MPTCP limits when flushing MPTCP
>>   endpoints. With a validation in the selftests. Fixes for v5.7.
>>
>> - Patches 4, 5, 6: Disallow '0' as ADD_ADDR retransmission timeout.
>>   With a preparation patch, and a validation in the selftests. Fixes for
>>   v5.11.
>>
>> - Patches 8, 9: Fix C23 extension warnings in the selftests, spotted by
>>   GCC. Fixes for v6.16.
> 
> userspace_pm.sh which hasn't flaked in 1000 runs has flaked last night,
> with this series applied:
> https://netdev-3.bots.linux.dev/vmksft-mptcp/results/255941/8-userspace-pm-sh/

Thank you for the notification!

> Looks unrelated but also quite strange?

Indeed: the whole series should not affect the behaviour validated by
the selftests. The error messages are not very useful, and the test was
not really slower than usual. Our CI never complained about that either.
I will monitor that!

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

Hello:

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

On Fri, 15 Aug 2025 19:28:18 +0200 you wrote:
> Here are various fixes:
> 
> - Patch 1: Better handling SKB extension allocation failures. A fix for
>   v5.7.
> 
> - Patches 2, 3: Avoid resetting MPTCP limits when flushing MPTCP
>   endpoints. With a validation in the selftests. Fixes for v5.7.
> 
> [...]

Here is the summary with links:
  - [net,1/8] mptcp: drop skb if MPTCP skb extension allocation fails
    https://git.kernel.org/netdev/net/c/ccab04469798
  - [net,2/8] mptcp: pm: kernel: flush: do not reset ADD_ADDR limit
    https://git.kernel.org/netdev/net/c/68fc0f4b0d25
  - [net,3/8] selftests: mptcp: pm: check flush doesn't reset limits
    https://git.kernel.org/netdev/net/c/452690be7de2
  - [net,4/8] mptcp: remove duplicate sk_reset_timer call
    https://git.kernel.org/netdev/net/c/5d13349472ac
  - [net,5/8] mptcp: disable add_addr retransmission when timeout is 0
    https://git.kernel.org/netdev/net/c/f5ce0714623c
  - [net,6/8] selftests: mptcp: disable add_addr retrans in endpoint_tests
    https://git.kernel.org/netdev/net/c/f92199f551e6
  - [net,7/8] selftests: mptcp: connect: fix C23 extension warning
    https://git.kernel.org/netdev/net/c/2eefbed30d46
  - [net,8/8] selftests: mptcp: sockopt: fix C23 extension warning
    https://git.kernel.org/netdev/net/c/3259889fd3c0

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html