Re: [PATCH mptcp-next v5 0/3] mptcp: add statistics for mptcp socket in use

[Matthieu Baerts <matthieu.baerts@tessares.net>]

Hi Menglong, Mat,

On 12/10/2022 02:53, Mat Martineau wrote:
> On Fri, 7 Oct 2022, menglong8.dong@gmail.com wrote:
> 
>> From: Menglong Dong <imagedong@tencent.com>
>>
>> In the 1th patch, we do some code cleanup with replease 'sock->sk'
>> with 'sk'. In the 2th patch, we add statistics for mptcp socket in
>> use. And in the 3th patch, we add the testing for this commit.
>>
> 
> I think v5 looks ok for the export branch - we can get more test
> coverage there and update if needed.
> 
> Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>

Thank you for the patch and the review!

I was going to apply it but I just noticed our public CI complained
about the new test during its second test:


 KVM Validation: debug:
  - Unstable: 2 failed test(s): selftest_diag selftest_mptcp_join 🔴:
  - Task: https://cirrus-ci.com/task/6538547548127232
  - Summary:
https://api.cirrus-ci.com/v1/artifact/task/6538547548127232/summary/summary.txt


The useful bit:

----------------------- 8< -----------------------
not ok 1 test: selftest_diag.tap # exit=7
# no msk on netns creation                          [  ok  ]
# listen match for dport 10000                      [  ok  ]
# listen match for sport 10000                      [  ok  ]
# listen match for saddr and sport                  [  ok  ]
# all listen sockets                                [  ok  ]
# after MPC handshake                               [  ok  ]
# ....chk remote_key                                [  ok  ]
# ....chk no fallback                               [  ok  ]
# msk in use statistics                             [  ok  ]
# msk in use statistics                             [  ok  ]
# check fallback                                    [  ok  ]
# many msk socket present                           [  ok  ]
# msk in use statistics                             [ fail ] expected
249 found 248
# msk in use statistics                             [  ok  ]
----------------------- 8< -----------------------

Do you mind looking at that please?


Please note that the public CI is known to be slow. Here it was failing
with a debug kernel config (KASAN, KMEMLEAK, etc. see
kernel/configs/debug.config file)

If you want to reproduce it in closer conditions, you can do something
like this to build the same kernel and run diag.sh in a loop until it fails:

  $ cd [kernel source code]
  $ echo "run_loop run_selftest_one ./diag.sh" > .virtme-exec-run
  $ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
      --pull always mptcp/mptcp-upstream-virtme-docker:latest \
      auto-debug

For more details:

    https://github.com/multipath-tcp/mptcp-upstream-virtme-docker


Cheers,
Matt
-- 
Tessares | Belgium | Hybrid Access Solutions
www.tessares.net