From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============7203983400440691838==" MIME-Version: 1.0 From: Geliang Tang To: mptcp at lists.01.org Subject: [MPTCP] Re: Weekly meeting - 25 June 2020 16:00 UTC (9am PDT, 6pm CEST) Date: Thu, 02 Jul 2020 13:31:16 +0800 Message-ID: <20200702053116.GA17066@OptiPlex> In-Reply-To: eeeb860d-601d-6533-1578-3c997de842f0@tessares.net X-Status: X-Keywords: X-UID: 4893 --===============7203983400440691838== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Matt, On Wed, Jul 01, 2020 at 06:17:14PM +0200, Matthieu Baerts wrote: > Hi Geliang, > = > On 26/06/2020 02:40, Geliang Tang wrote: > > Hi Matt, > > = > > There are several questions I didn't answer clearly in the last meeting. > > Here I try to write them down: > = > Thank you for the clarification! > = > > 1 My name pronunciation > > Geliang Tang sounds like Glenn Town. > = > Thank you, that's helpful! > = > > 2 My ideal meeting time > > Every hour from=C2=A011am to 6pm (my location time-zone) at every worki= ng day > > is fine to me to attend the meeting. > = > I just sent another email now that I am back to work. Did you see it? > I guess you subscribed to the ML but just in case: https://lists.01.org/h= yperkitty/list/mptcp(a)lists.01.org/thread/SHULDCJUWQJX7CWUH6N66MFTRW6TAODS/ > = > The best would be to have a single meeting, once per week where everybody > can join. But it would always be difficult for people of some time zones.= .. > I know it is very late and not in the working hours but do you think havi= ng > the meeting one hour sooner (11pm your time) would be OK for you? Maybe n= ot > every week of course. Don't be afraid to say no! > = 11pm is fine to me. Hope it is not too early to other people. -Geliang > An alternative would be to organise other meetings with less people. > But sadly, I don't see other possibilities. Feel free to share if any! > = > > 3 What would you=C2=A0like to work on? > > I'm interested in the BPF topic in the last meeting. But=C2=A0I don't h= ave > > any idea about this at the moment. Can you please give me more detail > > about it and I'll spend some time to learn it. > = > At the last meeting, we discussed about BPF because a new dev asked to wo= rk > on this subject. He is for the moment looking at having the possibility to > set in BPF some socket options per subflow. > = > Maybe later, he might look at adding BPF_CGROUP_SOCKET to allow per-cgroup > override of proto in the socket() call. More details in > https://github.com/multipath-tcp/mptcp_net-next/issues/18 > = > But still in BPF, it would be good to look at other stuffs like modifying > the behaviour of the Path Manager or the Scheduler with BPF. > Some works were done in a fork of the out-of-tree kernel, maybe it can he= lp: > https://dial.uclouvain.be/pr/boreal/object/boreal:225610 > = > Cheers, > Matt > -- = > Tessares | Belgium | Hybrid Access Solutions > www.tessares.net --===============7203983400440691838==--