Hi Paolo,
On 15/11/2022 18:45, Paolo Abeni wrote:
> The last bunch of mptcp fast open follow-up I have planned.
>
> This aims at reducing the number of unnecessary mptcp_crypto_key_sha()
> invocation, calling such primitive at most once per MPC subflow
> and caching the result as subflow level.
>
> The first patch should be inserted just before
>
> "mptcp: implement delayed seq generation for passive fastopen"
>
> while the 2nd one is a squash-to, still to the above.
>
> Note that this will cause a trivial conflict in "DO-NOT-MERGE: mptcp:
> use kmalloc on kasan build"
Thank you for the patches and the notes here above!
New patches for t/upstream:
- cccf62e31b88: mptcp: consolidate initial ack seq generation
- b4c09c888e0b: conflict in t/DO-NOT-MERGE-mptcp-use-kmalloc-on-kasan-build
- Results: dc870dc49415..ee79be9e1648 (export)
- 23c22cf366e6: "squashed" patch 2/2 in "mptcp: implement delayed seq
generation for passive fastopen"
- Results: ee79be9e1648..42fdee53dc8e (export)
Tests are now in progress:
https://cirrus-ci.com/github/multipath-tcp/mptcp_net-next/export/20221117T182326
Cheers,
Matt
--
Tessares | Belgium | Hybrid Access Solutions
www.tessares.net