Content-Type: multipart/mixed; boundary="===============0615494013518405155==" MIME-Version: 1.0 From: Matthieu Baerts To: mptcp at lists.01.org Subject: [MPTCP] Re: [PATCH v5 0/6] mptcp: just another complete datapath refactor Date: Thu, 19 Nov 2020 18:07:46 +0100 Message-ID: In-Reply-To: cover.1605789317.git.pabeni@redhat.com X-Status: X-Keywords: X-UID: 6835 --===============0615494013518405155== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Paolo, Florian, Mat, On 19/11/2020 13:38, Paolo Abeni wrote: > This iteration introduces some improvement for the wake-up/poll code > as discussed on the ML. > = > The last 2 patches in the previous iterations has been merged in a single > one, with all the relevant deltas noted there. > = > Looking for merge now ;) Done! Thank you for your patches and reviews! I just applied these patches in our tree after having fixed a few = checkpatch issues (only related to codespell), remove "." at the end of = the commit title, resolved one conflict with the patch (6/6) due to = "mptcp: update rtx timeout only if required." patch applied earlier = today and add Florian's Acked-by. @Mat: please tell me if you want me to add your Reviewed-by tags on = these ones ;-) - ea8e2035dfab: mptcp: open code mptcp variant for lock_sock - 8983b6ddc2a0: mptcp: implement wmem reservation - 44bb6c6d4e7f: mptcp: protect the rx path with the msk socket spinlock - 3a54a74a3c5b: mptcp: allocate TX skbs in msk context - 7ca128e90452: mptcp: avoid a few atomic ops in the rx path - 6b142355778a: mptcp: use mptcp release_cb for delayed tasks - Results: 55b2596bdf31..11818af71ea2 Tests + export are in progress. Note that I already updated the export branch manually = (export/20201119T175232) to let Christoph already starting some tests = with syzkaller. Cheers, Matt -- = Tessares | Belgium | Hybrid Access Solutions www.tessares.net --===============0615494013518405155==--